National Security Commission on Artificial Intelligence. Final Report - page 10
BLUEPRINT FOR ACTION: CHAPTER 15
o Given the urgent need to enhance the Department’s technology diplomacy
capacity and the likely long-term nature of the process of establishing a new Under
Secretary, efforts to implement this recommendation should proceed in parallel
with the Commission’s other organizational recommendations.
•
Establish a diplomatic presence in major U.S. and foreign technology hubs.
o The Department of State should enhance its presence in major foreign and U.S.
technology hubs, supported by establishing a cadre of dedicated technology
officers at U.S. missions to strengthen diplomatic advocacy, improve technology
scouting, and inform policy and foreign assistance choices.
o The Department should accelerate plans to establish 12 Regional Technology
Officer positions around the world72 and further describe how these officers will
enhance U.S. technology competitiveness with partners such as the Foreign
Commercial Service, USAID, and DFC. These officers should also scout technology
initiatives that can enhance our diplomatic and development capabilities.
o The Department should re-establish a permanent presence in Silicon Valley,
which it initiated in 2014, and established dedicated positions in 2015-2016. These
positions and State’s presence were discontinued when an OMB hiring freeze was
implemented in January 2017.
o In addition, the Diplomat in Residence program—with presence in 16 regions
at universities across America73—should be repurposed beyond recruitment to
include public diplomacy, technology scouting, and engagement with foreign
government and commercial entities active across America. Domestic insight is a
valuable input into foreign policy and will increase public confidence in and support
for America’s international technology leadership.
•
Incorporate AI and emerging technology training modules into Foreign Service
institute (FSI) courses.
o The Department of State should incorporate mandatory AI and emerging
technology-related modules into key FSI training courses, including the
Ambassadorial Seminar, the Deputy Chiefs of Mission course, Political and
Economic Tradecraf courses, and A-100 orientation training classes. FSI should
also develop a stand-alone course on emerging technologies and foreign policy.
o The Department should partner with academic and private-sector organizations to
access the leading edge of technology education while also building a more robust
technology fellows program for exchanges with industry.
Actions for Congress:
• Expedite necessary reorganization of the Department of State by passing
legislation to create an Under Secretary for Science, Research and Technology (Q).
o Congress should act to create the State/Q position and consolidate disparate
S&T efforts in the Department in a single division. There is urgent need for such
reorganization, and Congress can empower the Department of State by introducing
and passing legislation to expedite the reorientation.
• Appropriate funds to support immediate augmentation of the U.S. diplomatic
corps.
p
543
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
o Congress should provide robust funding for hiring and training of needed personnel
to enable the Department of State’s reorientation and support the Department’s
international efforts to promote U.S. values and standards in AI, data, and associated
emerging technologies.
o The Commission recommends that Congress provide at least $8 million in
supplemental appropriations for immediate hiring of staff to address emerging
technology needs across the Department’s offices and bureaus, to establish a
diplomatic presence in major U.S. and foreign technology hubs, and to develop FSI
courses.
o See the Funding Table Appendix to this report for a detailed breakdown of the
proposed appropriations.
•
Appropriate funds to support the CSET Bureau.
o The Commission recommends a minimum of $20 million to establish the CSET
Bureau.
o See the Funding Table Appendix to this report for a detailed breakdown of the
proposed appropriations.
•
Appropriate funds to support critical diplomatic efforts, programs, and foreign
operations in AI, emerging technologies, and data.
o Further funding is needed to enable the Department of State to advance
responsible AI aligned with U.S. and like-minded values.
o While details of funding needs should reflect input from the Department of State,
the Commission recommends, at a minimum, that Congress issue a supplemental
appropriation for no less than $37 million, as a subset of the proposed $300 million
described earlier in this Blueprint for Action, for the following urgent needs:
■ Public diplomacy messaging and engagement to support democratic values
and raise awareness of U.S. leadership in AI innovation as well as the risks of
unwanted technology transfer and authoritarian digital practices;
■ AI exchange programs to promote U.S. values and fund participation by
developing countries participation in multilateral AI activities;
■ Programs to showcase American innovation and through which to promote the
ethical use of AI, including the American Spaces, TechCamp, MakerSpaces,
and U.S. Speakers programs74;
■ Partnership development and advancement of scientific norms through the
U.S. Science Envoys and Embassy Science Fellows programs;
■ Diplomatic operations and programs around international AI cooperation,
including support for initiatives of the ETC;
■ Promotion of human rights and fundamental freedoms in the AI context;
■ Maintaining U.S. lead in the use of AI for military applications through
cooperation with allies and partners;
■ Ensure political and policy congruence with allies and partners on the use of
AI-enabled weapon systems;
p
544
BLUEPRINT FOR ACTION: CHAPTER 15
■ Ensure continued interoperability among the U.S. and its allies and partners;
■ Training and capacity-building for foreign governments on emerging
technologies to support responsible innovation;
■ Reporting to counter malign influence in AI ecosystems;
■ Empower global AI-focused S&T entrepreneurship through the U.S. Global
Innovation through Science and Technology (GIST) Initiative75; and
■ Public diplomacy initiatives on international AI standards as well as tracking
and reporting on public opinion related to AI.
■ See the Funding Table Appendix to this report for a detailed breakdown of the
proposed appropriations.
Table 1. Key Multilateral Technology Initiatives
Initiatives
Critical Areas
Objectives
Council of
•
Standards and Norms
•
Support the Ad Hoc Committee on Artificial Intelligence (CAHAI)
Europe/CAHAI
•
Promote Human Rights
efforts around the development and use of AI aligned with human
and Democracy
rights, rule of law, and democracy (goal of international legislative
framework on AI similar to the Budapest Convention on Cybercrime)
•
Membership: 47 countries from Europe and six Observer States
(Canada, US, Israel, Japan, Mexico, Holy See)
D10 Initiative
•
Standards and Norms
•
Foster international cooperation to provide 5G alternatives to ZTE and
•
Promote Human Rights
Huawei; shift critical supply chains out of China, and protect national
and Democracy
security
•
Support nascent effort as it builds on the promising coalition and
refines its goals, structure, and timeline; explore potential for the UK
and U.S. to jointly announce further developments
•
Explore expansion into other emerging technologies critical to
national security
•
Additional consideration: Divergent views on 5G and absence of key
nations in 5G effort may limit efficacy
•
Membership: Australia, Canada, France, Germany, India, Italy, Japan,
South Korea, the UK (founding member), and the U.S.
OECD Digital
•
Standards and Norms
•
Continue articulating support of OECD’s international efforts to
and AI efforts
•
Data Sharing
advance responsible AI and promote implementation of OECD AI
(OECD AI
Principles
•
IDDI
Principles,
•
Advance shared interests through AI Policy Observatory (OECD.AI)
•
Joint R&D on AI and
OECD.AI,
and the Network of Experts on AI (ONE AI), particularly in working
Digital Infrastructure
groups on classifying AI systems, implementing values-based
ONE AI,
principles, and guiding national AI strategies
“Going Digital”
•
Facilitate coordination of U.S. experts and representatives engaging
Project)
in ONE AI and associated working groups
•
Advance U.S. interests in “Going Digital” initiatives that promote data
sharing and harmonizing on IP and regulation
•
Membership: 37 member countries, OECD AI Principles adopted by
42 countries
p
545
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Initiatives
Critical Areas
Objectives
Freedom
•
Standards and Norms
•
Engage to further efforts on AI and human rights and against digital
Online
•
IDDI
authoritarianism, with a focus on content moderation and facial
Coalition
recognition
(FOC)/ T-FAIR
•
Support Task Force on AI and Human Rights (T-FAIR)
•
Membership: 32 country coalition with members from Africa to Asia,
Europe, the Americas, and the Middle East
GPAI
•
Standards and Norms
•
Influence direction, scope, and goals; support expanded membership
•
Data Sharing
•
Formalize a U.S.-based center of expertise to provide technical
•
Joint R&D on AI and
support for working groups
Digital Infrastructure
•
Advance shared interests of democratic nations through working
groups:
1) Responsible AI
(including Ad Hoc AI and Pandemic
•
IDDI
Response Subgroup), 2) Data governance, 3) Future of work, 4)
Commercialization and Innovation and facilitate coordination of U.S.
experts and representatives engaging in working groups and steering
committee
•
Engage in Multi-stakeholder Experts Group Plenary
•
Membership: Australia, Canada, France, Germany, India, Italy, Japan,
Mexico, New Zealand, Singapore, Slovenia, South Korea, Spain, the
United Kingdom, the United States, and the European Union with the
OECD and UNESCO as a Permanent Observer the GPAI Council and
Steering Committee
IPAC
•
Promote Human Rights
•
Influence IPAC to become a vehicle for promoting AI-related goals
and Democracy
of democratic nations, including the spread of democratic technology
alternatives; alignment on emerging technologies that pose threats
•
Standards and Norms
to national security; alignment on fair trade practices for digital
•
Promote and Protect
commerce; countering IP theft; alignment on export controls and
Innovation
investment screening; and overall legislative alignment
•
IDDI
•
Membership: IPAC Co-chairs span the political spectrum and come
•
Defense and Security*
from 19 countries across North America, Europe, Asia, and Africa
Five Eyes/
•
Defense and Security
•
Develop methods to address AI application and interoperability,
TTCP AI
including a possible test bed for application in other situations and
Strategic
with other coalitions (e.g., NATO)
Challenge
•
Membership: Australia, Canada, New Zealand, the United States, and
the United Kingdom
JAIC AI
•
Defense and Security
•
Strengthen role as key multilateral forum for security coordination
Partnership
•
Data Sharing
around AI, including advancing shared interests and best practices on
for Defense
AI ethics, collaborative frameworks, and strategic messaging
•
Standards and Norms
•
Current membership: United States plus twelve partner nations—
Australia, Canada, Denmark, Estonia, Finland, France, Israel, Japan,
Norway, South Korea, Sweden, and the United Kingdom
NATO
•
Defense and Security
•
Promote interoperability, human capital development, implementation
of strategic objectives
•
Data Sharing
•
Standards and Norms
•
Membership: 30 countries from Europe and North America
National
•
Defense and Security
•
Leverage NTIB to strengthen the industrial capabilities of the U.S.
Technology
and allies and address supply chain concerns; explore expansion of
•
Promote and Protect
and Industry
Innovation
mandate to advance AI-related interests around defense and security
issues
Base (NTIB)
•
Membership: Australia, Canada, the United Kingdom, and the United
States
*In
addition to mapping multilateral efforts on
AI and associated technologies to the seven critical areas
recommended as priorities for the Emerging Technology Coalition, and detailed in the Annex: Emerging
Technology Coalition Blueprint, this table includes multilateral efforts that are important for defense and security,
particularly as articulated in Chapter 3: AI-Enabled Warfare.
p
546
BLUEPRINT FOR ACTION: CHAPTER 15
Initiatives
Critical Areas
Objectives
The Quad
•
Defense and Security
•
Build on the Quad framework to deepen AI cooperation and negotiate
formal AI cooperation agreements in the Indo-Pacific
•
Membership: Australia, India, Japan, the United States
Wassenaar
•
Defense and Security
•
Advance multilateral coordination on export controls on conventional
Arrangement
•
Promote and Protect
and dual-use technologies
Innovation
•
Membership: 42 participating states—Australia, Argentina, Canada,
•
Promote Human Rights
India, Japan, Mexico, New Zealand, Russia, South Africa, South
and Democracy
Korea, Turkey, Ukraine, the United Kingdom, the United States, and
all EU members other than Cyprus
G7
•
Standards and Norms
•
Support French-led effort to use G7 to promote responsible AI
•
Data Sharing
development and coordinate on efforts to counter disinformation and
other dangerous online content
•
Promote and Protect
Innovation
•
Additional consideration: Success of D10 as a coalition to address
common AI-related issues may limit G7 efficacy in the area, although
it continues to serve as a key forum to address important geopolitical
topics
•
Membership: Canada, France, Germany, Italy, Japan, the United
Kingdom, and the United States
G20
•
Joint R&D on AI and
•
Advance efforts to enable international digital economy, develop
Digital Infrastructure
global solution to tax challenges from digitization of the economy, and
•
Standards and Norms
utilize technology in infrastructure and smart city efforts
•
Data Sharing
•
Ensure countries do not successfully promote authoritarian technology,
particularly on topic of smart cities
•
Membership: Governments of 19 countries and the EU; includes
China, Russia
IP5
•
Promote and Protect
•
Continue to engage in IP5’s New Emerging Technologies AI Task
Innovation
Force to advance global legal certainty and protections of AI-related
IP, enhance efficiencies in office operations through AI adoption, and
strengthen communication with industry
•
Membership: European Patent Office and national patent offices from
Japan, South Korea, China, and the United States
3GPP
•
Standards and Norms
•
Advance standards to enable innovation, protect national and
economic security in telecommunications
•
Promote and Protect
Innovation
•
Defend the integrity of technical standardization
•
Additional considerations: Chinese companies participating as voting
members have more than doubled in the past few years to 110 in
January 2020, more than twice the 53 US voting members
•
Membership: Seven telecommunications standards development
organizations from Japan, the United States, China, Europe, and
South Korea
IEEE
•
Standards and Norms
•
Engage, particularly on standards within the P7000 series on ethically
•
Data Sharing
aligned design series (e.g., P7001 - Transparency of autonomous
systems and P7003 - Algorithmic Bias) and Ethics Certification
Program for Autonomous and Intelligent Systems (ECPAIS)
•
Membership: IEEE is a technical professional organization that has
419,000 members from more than 160 countries
ISO/IEC
•
Standards and Norms
•
Advance standards to enable innovation; protect national and
economic security
•
Data Sharing
•
Maintain consensus approach to standards development; counter
adversarial or politicization efforts
•
Ensure U.S. domestic policy and resourcing enables full U.S.
engagement, particularly in ISO/IEC JTC 1/SC 42 on AI
•
Membership: 31 participating member countries and 16 observing
member countries/territories
p
547
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Initiatives
Critical Areas
Objectives
ITU-T
•
Standards and Norms
•
Continue to engage on international technical standards and defend
the integrity of international technical standards
•
Data Sharing
•
Engage in AI for Global Good Summit, which has strong involvement
from India, China, and Japan
•
Membership: 193 countries are members of ITU-T
UNESCO
•
Standards and Norms
•
Engage with the Ad Hoc Expert Group’s initiative to create a global
standard-setting instrument on ethics and AI (Recommendation on
Ethics of AI)
•
Support continued development of the AI Decision Makers’ Essential
Toolkit to help decision makers, particularly in Africa, address practical
questions for AI and UNESCO’s AI capacity building programs for
stakeholders in the judicial system
•
Membership: 193 member countries
UN CCW
•
Standards and Norms
•
Advance shared interests of democratic nations regarding lethal
GGE on LAWs
autonomous weapon systems
•
Membership: Over 80 countries have participated in discussions
UNSG High-
•
Standards and Norms
•
Engage as part of UN engagement; however, it is unlikely the High-
Level Panel
•
Promote Human Rights
level Panel on Digital Cooperation (HLPDC) will be a main vehicle for
on Digital
and Democracy
advancing U.S. AI interests
Cooperation
•
Membership: Panel has participants from over 15 countries
WIPO AI
•
Promote and Protect
•
Continue to engage in WIPO’s “Conversations” on AI and IP Policy
and IP
Innovation
and Administration; includes data protection and sharing standards
Conversations
•
Data Sharing
•
Membership:
193 member countries, with participants in the
conversations on IP and AI receiving participants from over 130
countries
WTO
•
Promote and Protect
•
Continue to engage in e-commerce and trade efforts outflowing
Innovation
from the 2019 Joint Statement on Electronic Commerce and the
•
Data Sharing
“Osaka Track,” which promotes international rule-making to promote
e-commerce and addresses data concerns
•
Membership: Although there are 164 member countries in WTO, 78
members signed the Joint Statement on Electronic Commerce; 24
countries signed the Osaka Declaration on the Digital Economy
p
548
BLUEPRINT FOR ACTION: CHAPTER 15
Table 2: International Digital Development Programs
Trilateral Infrastructure Partnership (TIP). Established in November 2018 by the US, Japan, and Australia, the TIP supports
infrastructure projects for “an Indo-Pacific region that is free, open, inclusive, prosperous, and secure.” TIP coordinates
separate financing and investment efforts of Australian Infrastructure Financing Facility for the Pacific, Japan’s Bank for
International Cooperation (JBIC), Japan’s Nippon Export and Investment Insurance, and U.S. government stakeholders
(DFC, USAID, Department of the Interior Compact Funding, and the U.S. government’s Transaction Advisory Fund). In
October 2020, TIP facilitated a $30 million project to construct underseas fiber optic cable to Palau to ensure secure digital
connectivity.
Blue Dot Network (BDN). Launched in November 2019 by the US, Japan, and Australia, BDN convenes “governments, the
private sector, and civil society to promote high-quality, trusted standards for global infrastructure development in an open
and inclusive framework.” It leverages “commonly accepted principles and standards to promote market-driven, transparent,
and financially sustainable infrastructure development” in the Indo-Pacific region and globally. It is led by the U.S. DFC,
Australia’s Department of Foreign Affairs and Trade, and the JBIC.
World Bank - Digital Development Partnership (DDP). Established in 2016 to support the implementation of the UN’s
Sustainable Development Goals, the DDP has over 50 active client countries with an emphasis on delivering data and
indicators, digital economy enabling environment, cybersecurity, accessible internet, digital government, and mainstreaming
digital services, solutions, and platforms.
Three Seas Initiative Investment Fund (3SIIF). The 3SIIF is a partnership of 12 countries focused on promoting economic
growth, security, and a more cohesive and stronger Europe through infrastructure development in the energy, transportation,
and digital sectors. The Three Seas Initiative Investment Fund (3SIIF) (currently over $900 million, including a $300 million
DFC initial investment) supports digital infrastructure projects to meet regional compute, storage, and connectivity needs.
EU Economic and Investment Plan for the Western Balkans. Announced in October 2020, this plan supports the region’s
digital and green transitions while boosting economic development, regional cooperation, and EU integration. Through
investment of 9 billion euros and the Western Balkans Guarantee Facility, which seeks to mobilize 20 billion euros in
public and private investment, it aims to bolster digitalization across ten investment flagships. Projects include broadband
infrastructure in six Western Balkan partners, establishing trustworthy data centers and cloud infrastructures, expanding the
Balkan Digital highway, and digital education programs.
EU-Asia Connectivity Strategy. In 2018, the European Parliament, the European Council, the European Economic and
Social Committee, the Committee of the Regions and the European Investment Bank issued a joint communication to
establish building blocks on an “EU Strategy on Connecting Europe and Asia” through interoperable transport, energy and
digital networks. Efforts are currently underway to strengthen EU-Asia connectivity that is sustainable, comprehensive and
adherent to the international rules-based order.
FAIR Forward - Artificial Intelligence for All. The German Development Cooperation initiative seeks to create a more
“open, inclusive and sustainable approach to AI,” particularly in Africa and Asia. In partnership with Ghana, Rwanda, South
Africa, Uganda, and India, FAIR Forward seeks to: 1) strengthen local technical AI know-how, 2) remove entry barriers to
AI through access to training data and AI technologies, and 3) develop policy frameworks for ethical AI, data protection,
and privacy. Projects include an African AI start-up accelerator, a Pan-African digitalization initiative, and the Open for Good
Alliance to improve localized AI training data.
International Digital Cooperation - ICT Standardization (InDiCo). Launched in 2018 by the European Commission,
InDiCo is a three-year project to promote ICT standards alignment and interoperability with key partner countries, including
Japan, South Korea, China, India, Brazil, and the United States. It coordinates technical standardization meetings and
organizes technical and political workshops on ICT standards.
Quality Infrastructure Initiative. Japan has championed efforts to promote “quality infrastructure” investments, including
through financing of approximately $200 billion. Japan’s efforts have led to the development of principles for infrastructure
investments and is expanding into digital connectivity.
Global Cooperation and Training Framework (GCTF). GCTF, led by Taiwan and in partnership with the United States and
Japan, serves as a platform for Taiwan to share its expertise with partners around the world. Recently, GCTF held a virtual
webinar with Latin American and Caribbean governments on digitization, particularly on ways to leverage data and AI to help
governments respond to COVID-19.
p
549
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Fact Sheet, U.S. Department of State, The United States Partners with Australia and Japan to Expand
Reliable and Secure Digital Connectivity in Palau (Oct.
29,
united-states-partners-with-australia-and-japan-to-expand-reliable-and-secure-digital-connectivity-
in-palau//index.html; Fact Sheet, U.S. Department of State,
2020 Indo-Pacific Business Forum
business-forum-promotes-free-and-open-indo-pacific/index.html; The Launch of Multi-Stakeholder
dfc.gov/media/opic-press-releases/launch-multi-stakeholder-blue-dot-network; Blue Dot Network,
U.S. Department of State (last accessed Feb. 16, 2021), https://www.state.gov/blue-dot-network/; Blue
Dot Network: Frequently Asked Questions, U.S. Department of State (last accessed Feb. 16, 2021),
https://www.state.gov/blue-dot-network-frequently-asked-questions/; Digital Development Partnership
development-partnership; Digital Development Partnership Annual Report: Responding to the COVID-19
Crisis, The World Bank
(Oct.
26,
digital-development-partnership-annual-report-responding-to-the-covid-19-crisis; Objectives, Three
Seas Initiative
(last accessed Feb. 16, 2021), https://3seas.eu/about/objectives; Three Seas Story,
Three Seas Initiative (last accessed Feb. 16, 2021), https://3seas.eu/about/threeseasstory; The Three
Seas Fund Makes Its First Digital Investment, Three Seas Initiative Investment Fund (Dec. 2, 2020),
https://3siif.eu/news/the-three-seas-fund-makes-its-first-digital-investment; The Fund, Three Seas
Initiative Investment Fund (last accessed Feb. 16, 2021), https://3siif.eu/the-fund/; Press Release, U.S.
International Development Finance Corporation, DFC Approves Over $2.1 Billion in New Investments
for Global Development
(Dec.10,
over-21-billion-new-investments-global-development; Press Release, European Commission, Western
Balkans: An Economic and Investment Plan to Support The Economic Recovery and Convergence
(Oct.
6,
2020), https://ec.europa.eu/commission/presscorner/detail/en/ip_20_1811; Questions and
Answers: Economic and Investment Plan for the Western Balkans, European Commission (Oct. 6,
2020), https://ec.europa.eu/commission/presscorner/detail/en/qanda_20_1819; Connecting Europe
europa.eu/sites/eeas/files/joint_communication_-_connecting_europe_and_asia_-_building_blocks_
for_an_eu_strategy_2018-09-19.pdf; Connecting Europe & Asia: The EU Strategy, European External
Action Service (Sept. 26, 2019), https://eeas.europa.eu/headquarters/headquarters-homepage/50699/
node/50699_en; FAIR Forward - Artificial Intelligence for All, Deutsche Gesellschaft für Internationale
Zusammenarbeit
(GIZ) GmbH
(Jun.
www.openforgood.info/; About the Project, The InDiCo Project (last accessed Feb. 16, 2021), https://
www.indico-ictstandards.eu/about-the-project; The ‘Expanded Partnership for Quality Infrastructure’
Initiative Directed Toward the G7 Ise-Shima Summit Meeting Announced, Japanese Ministry of Economy,
Trade and Industry (May 23, 2020), https://www.meti.go.jp/english/press/2016/0523_01.html; Tobias
Harris, ‘Quality Infrastructure’: Japan’s Robust Challenge to China’s Belt and Road, War on the Rocks
(Apr.
9,
chinas-belt-and-road/; Andreea Brînză, Japan’s Belt and Road Balancing Act, The Diplomat (Nov. 8,
2018), https://thediplomat.com/2018/11/japans-belt-and-road-balancing-act/; Global Cooperation and
Training Framework (GCTF) Programs, American Institute in Taiwan (last accessed Dec. 28, 2020),
https://www.ait.org.tw/our-relationship/global-cooperation-and-training-framework-programs-gctf/;
Fact Sheet, U.S. Department of State, U.S. Support for Digital Transformations in Latin America and the
Caribbean (Nov. 10, 2020), https://2017-2021.state.gov/u-s-support-for-digital-transformation-in-latin-
america-and-the-caribbean//index.html.
p
550
BLUEPRINT FOR ACTION: CHAPTER 15
Table 3: U.S. Digital Development Programs
Infrastructure Transaction and Assistance Network (ITAN) and Transaction Advisory Fund (TAF). ITAN and TAF are
multi-agency, United States Government efforts to strengthen capacity-building programs, provide expert transaction advisory
services, and coordinate U.S. assistance support for sustainable, transparent, high-quality digital and non-digital infrastructure
in the Indo-Pacific region.
US-ASEAN Smart Cities Partnership (USASCP). Launched in 2018, USASCP aims to promote U.S. private sector
engagement in smart, sustainable city solutions, sharing best practice and technical collaboration, and strengthening digital
economy and cybersecurity. It draws on capabilities of the Departments of State, Commerce, and Transportation as well as
USAID, USTDA, and the NSF. Limited funding ($10 million initial investment in 2018) has limited its impact.
Digital Connectivity and Cybersecurity Partnership (DCCP). DCCP draws on tools of 10 U.S. agencies to promote
expanded connectivity and an open, interoperable, secure, and reliable internet. In the Indo-Pacific, $50 million has supported
15 countries through capacity building to improve digital connectivity; strengthen the private sector’s digital capacity; and assist
in the design, development, and implementation of ICT policies and regulations. In the Western Hemisphere, $10 million will
support similar initiatives.
Bilateral Memorandums of Understanding and Investments. United States Government entities are undertaking various
bilateral projects to advance a democratic digital ecosystem. For example, EXIM and the Brazilian Ministry of Economy signed
a MOU to deploy up to $1 billion in financing to support U.S. exports to Brazil in the 5G space. DFC is financing a major
telecommunications project in Ecuador and Peru, which will deploy at least 500 telecom towers and expand access to 4G
mobile broadband and high-speed internet in rural areas.
USAID’s Digital Strategy. USAID has a far-reaching plan for digital capacity building and foreign assistance, especially around
strengthening open, secure, and inclusive digital ecosystems; providing country assessments to inform digital development;
and supporting the responsible use of digital technologies. For example, USAID will support the African Union’s drafting of a
dedicated protocol on digital trade and e-Commerce within the African Continental Free Trade Area. However, global demand
significantly exceeds available funding for the digital programs. In FY2019, for example, USAID provided only $1.7 million in
foreign assistance under the Digital Ecosystem Fund, one component of the strategy.
DFC’s Roadmap for Impact. The Roadmap proposes catalyzing $75 billion—$25 billion in DFC funds plus $50 billion from
the private sectors—over five years (2020-2025) to support development projects in low- and lower middle-income countries.
DFC’s investment focus on “Technology and Infrastructure” includes support for open, interoperable, reliable and secure digital
infrastructure and internet access. The DFC five-year plan aspires to commit $5 billion in technology and critical infrastructure
investments through 10 major infrastructure projects.
EXIM’s Program on China and Transformational Exports (PCTE). Established by Congress in December 2019, PCTE
directs EXIM to reserve no less than 20% of its total financing authority ($27 billion out of $135 billion) “to directly neutralize
export subsidies for competing goods and services financed by official export credit, tied aid, or blended financing provided
by China or by other covered countries” and “to advance the comparative leadership of the U.S. with respect to China, or
support United States innovation, employment, and technological standards, through direct exports” in specific industries. AI
and associated technologies are an explicit focus.
USTDA’s Global Infrastructure Resilience Initiative. The Initiative supports work with emerging markets to plan, sustain,
and finance infrastructure to combat external threats. Projects include ICT, remote health technology, emergency response
technology, and grid infrastructure technology. USTDA also collaborates with USAID and other agencies to build digital
infrastructure and provide technical assistance as part of the DCCP.
p
551
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Advancing Sustainable Infrastructure in the Indo-Pacific Region, USAID
(last accessed Feb.
16,
2021),
pdf; Fact Sheet, U.S. Department of State, U.S.-ASEAN Smart Cities Partnership (USASCEP): Sharing
s-asean-smart-cities-partnership-usascp-sharing-expertise-between-cities-to-benefit-the-people-of-
usascp.org/home-page; Digital Connectivity and Cybersecurity Partnership (DCCP), USAID (Oct. 19,
2020),
https://www.usaid.gov/digital-development/digital-connectivity-cybersecurity-partnership;
Fact Sheet, U.S. Embassy and Consulate in the Republic of Korea, 2020 Indo-Pacific Business Forum
pacific-business-forum-promotes-free-and-open-indo-pacific/; Press Release, U.S. Embassy Chile,
U.S. Support for Digital Transformation in Latin America and the Caribbean (Nov. 10, 2020), https://
cl.usembassy.gov/u-s-support-for-digital-transformation-in-latin-america-and-the-caribbean/;
Fact
Sheet, U.S. Department of State, U.S. Support for Digital Transformations in Latin America and the
Caribbean (Nov. 10, 2020), https://2017-2021.state.gov/u-s-support-for-digital-transformation-in-latin-
america-and-the-caribbean//index.html; USAID has published resources, such as Reflecting the Past,
Shaping the Future: Making AI Work for International Development and Managing Machine Learning
Projects in International Development: A Practical Guide, which are reflective of the Digital Strategy
and outline democratic principles in the deployment of those technologies. See Reflecting the Past,
Shaping the Future: Making AI Work for International Development, USAID (May 21, 2019), https://www.
usaid.gov/digital-development/machine-learning/AI-ML-in-development; Managing Machine Learning
Projects in International Development: A Practical Guide, USAID (Jan. 12, 2021), https://www.usaid.gov/
digital-development/managing-machine-learning-projects; Digital Strategy 2020-2024, USAID, (Jun.
2020), https://www.usaid.gov/usaid-digital-strategy; Digital Ecosystem Fund: 2020 Activities, USAID
(Dec. 22, 2020), https://www.usaid.gov/digital-development/DEF2020; Roadmap for Impact, DFC (last
accessed Feb. 16, 2021), https://www.dfc.gov/roadmap-for-impact; Fact Sheet, Export-Import Bank
of the United States, Overview: Program on China and Transformational Exports (last accessed Feb.
16,
exports-program/fact-sheet#:~:text=EXIM%20is%20actively%20working%20to,or%20by%20other%20
covered%20countries; Fact Sheet, Export-Import Bank of the United States, Program on China and
Transformational Exports: Supporting Artificial Intelligence (last accessed Feb. 16, 2021), https://www.
exim.gov/who-we-serve/external-engagement/supporting-us-artificial-intel-exports; Press Release,
Export-Import Bank of the United States, EXIM Board Unanimously Approves Historic Policy to Support
gov/news/exim-board-unanimously-approves-historic-policy-support-exporters-competing-peoples-
republic; Global Infrastructure Resilience Initiative, U.S. Trade and Development Agency (last accessed
Feb.
16,
2020), https://ustda.gov/call-for-initial-proposals/;
2021 Indo-Pacific Strategy Fact Sheet,
U.S. Trade and Development Agency (last accessed Feb. 16, 2020), https://ustda.gov/wp-content/
uploads/2021-Indo-Pacific-Strategy-Fact-Sheet.pdf; Press Release, U.S. Trade and Development
Agency, USTDA Announces 2021 Digital Strategy for the Indo-Pacific (Oct. 27, 2020), https://ustda.gov/
ustda-announces-2021-digital-strategy-for-the-indo-pacific/.
p
552
BLUEPRINT FOR ACTION: CHAPTER 15
Blueprint for Action: Chapter 15 - Endnotes
1 For more information on the National Technology Strategy, see Chapter 9 of this report.
2 See Chapter 9 of this report for more details.
www.nscai.gov/previous-reports/.
4 The Commission supports these efforts and further encourages the U.S. government to engage
proactively through a “mosaic approach” to ensure the Emerging Technology Coalition is additive and
not duplicative. See Interim Report and Third Quarter Recommendations, NSCAI at 185 (Oct. 2020),
https://www.nscai.gov/previous-reports/.
5 Key organizations for membership include the OECD, the Council of Europe (through the Ad Hoc
Committee on AI), the Freedom Online Coalition, GPAI, and the North Atlantic Treaty Organization.
The United States and other core partner states should consider including international organizations
as observers (e.g., the World Bank, the International Monetary Fund, the World Intellectual Property
Organization, the World Health Organization, the World Trade Organization, and United Nations
Educational, Scientific, and Cultural Organization).
legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449.
7 The IDDI should mobilize financial resources and technical expertise as the DFI Alliance, a
partnership between the U.S. International Development Finance Corporation and development
finance institutions (DFIs) of 15 other OECD countries, ushered in response to COVID-19. See
Development Finance Institutions Join Forces to Respond to COVID-19 in Developing Countries, U.S.
International Development Finance Corporation (April 6, 2020), https://www.dfc.gov/media/press-
releases/development-finance-institutions-join-forces-respond-covid-19-developing.
8 The Global Infrastructure Hub has forecasted global telecommunications infrastructure investment
need at $8.9 trillion over the next approximately 20 years, with current trends falling short of the
need by $1 trillion. Forecasting Infrastructure Investment Needs and Gaps, Global Infrastructure
Hub (2020), https://outlook.gihub.org/. The Alliance for Affordable Internet estimates it will cost
$428 billion and up to 10 years to achieve universal connectivity to quality broadband internet. See
Maiko Nakagaki, $428 Billion Investment Needed to Connect All of Humanity to the Internet by 2030,
Alliance for Affordable Internet (Sept. 17, 2020), https://a4ai.org/428-billion-investment-needed-to-
connect-all-of-humanity-to-the-internet-by-2030/.
9 This could include bilateral development finance institutions (DFIs) in OECD member countries and
multilateral DFIs, which are the private-sector arms of multi-state IFIs. See Development Finance
org/development/development-finance-institutions-private-sector-development.htm.
10 See the Chapter 7 Blueprint for Action and the Annex of this report containing the abridged version
of NSCAI’s Key Considerations for Responsible Development & Fielding of AI. For additional details
on the Commission’s recommendation to employ technologies and operational policies that align with
privacy preservation, see the section on “Aligning Systems and Uses with American Values and the
Rule of Law” in Key Considerations for Responsible Development & Fielding of Artificial Intelligence:
Extended Version, NSCAI (2021) (on file with the Commission).
11 G20 Principles for Quality Infrastructure Investment (last accessed Jan. 4, 2021), https://www.mof.
go.jp/english/international_policy/convention/g20/annex6_1.pdf.
12 “The Digital Principles were first created in consultation with organizations such as The Bill
and Melinda Gates Foundation, the Swedish International Development Agency (SIDA), the UN’s
Children’s Fund (UNICEF), UN Development Program (UNDP), the World Bank, and the U.S.
Agency for International Development (USAID), and the World Health Organization (WHO).” See
Frequently Asked Questions, Principles for Digital Development (last accessed Jan. 3, 2021), https://
digitalprinciples.org/about/.
13 Criteria for Security and Trust in Telecommunications Networks and Services, CSIS Working Group
on Trust and Security in 5G Networks (May 2020), https://csis-website-prod.s3.amazonaws.com/s3fs-
public/publication/200511_Lewis_5G_v3.pdf.
p
553
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
14 Recommendation of the Council on Digital Security of Critical Activities, OECD (Oct. 12, 2019),
https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0456.
mne/48004323.pdf.
16 Guiding Principles on Business and Human Rights: Implementing the United Nation “Protect,
Respect and Remedy” Framework, United Nations Human Rights Office of the High Commissioner
(2011), https://www.ohchr.org/Documents/Publications/GuidingPrinciplesBusinessHR_EN.pdf. To
mark the 10th anniversary of the adoption of the UN Guiding Principles on Business & Human Rights
(UNGPs), the UN will review existing gaps and develop a roadmap for the next decade. See UN
Guiding Principles: The Next Decade, Business & Human Rights Resource Centre (last accessed Feb.
human-rights/un-guiding-principles-the-next-decade/.
financing-sustainable-development/development-finance-standards/; DAC and CRS Code Lists,
OECD (last accessed Dec. 28, 2020), http://www.oecd.org/dac/financing-sustainable-development/
development-finance-standards/dacandcrscodelists.htm.
legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449.
19 See the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details, see the Key Considerations for
Responsible Development & Fielding of Artificial Intelligence: Extended Version, NSCAI (2021) (on file
with the Commission).
20 See Chapter 14 of this report for additional details and recommendations pertaining to end-use and
end-user export controls.
21 For example, Hillhouse Capital, an Asia-focused private equity firm known for its early investments
in Tencent and Baidu, grew from a “boutique hedge fund into a $60 billion behemoth that’s made
prescient bets on stocks, private equity and venture capital.” Hillhouse currently seeks to raise “what
would be Asia’s largest U.S. dollar-denominated fund targeting $13 billion.” See Michael McDonald
& Lulu Yilun Chen, Hillhouse Reloads After Building $60 Billion Asia Juggernaut, Bloomberg (April
hillhouse-ready-to-reload; Kane Wu & Julie Zhu, Exclusive: Hillhouse Targets Over $3 Billion for New
hillhouse-fundraising-exclusive/exclusive-hillhouse-targets-over-3-billion-for-new-yuan-denominated-
fund-sources-idUSKBN2690LK.
22 See Chapters 11, 13, 14, and 16 of this report, along with their associated Blueprints for Action,
for recommendations to strengthen public-private partnerships and private-sector investments in the
United States.
gov/previous-reports/.
24 This executive order would build upon Executive Order 13859. Donald J. Trump, Executive Order on
Maintaining American Leadership in Artificial Intelligence, The White House (Feb. 11, 2019), https://
trumpwhitehouse.archives.gov/presidential-actions/executive-order-maintaining-american-leadership-
artificial-intelligence/; see also Interim Report and Third Quarter Recommendations, NSCAI at 207-12
(Oct. 2020), https://www.nscai.gov/previous-reports/.
gov/previous-reports/.
26 See the Chapter 15 Blueprint for Action Annex. For additional recommendations on how NIST can
support qualified confidence in AI models and predicted outcomes, see Chapter 7 of this report and
its associated Blueprint for Action.
p
554
BLUEPRINT FOR ACTION: CHAPTER 15
Blueprint for Action: Chapter 15 - Endnotes
27 This includes the American National Standards Institute, the primary industry organization
advocating for U.S. companies before international standards bodies, and the International
Digital Economy and Telecommunication Advisory Committee (IDET). See Interim Report and
Third Quarter Recommendations, NSCAI at 208-209 (Oct. 2020); About IDET, U.S. Department
of State (last accessed Feb. 11, 2021), https://www.state.gov/international-digital-economy-and-
telecommunication-advisory-committee/about-idet/.
gov/previous-reports/.
29 Id.
30 Id.
31 The Task Force should also evaluate efforts to promote targeted development priorities. For
example, Power Africa is a “U.S. Government-led partnership, coordinated by USAID, that brings
together the collective resources of over 150 public- and private-sector partners to double access to
sites/default/files/documents/power-africa-fact-sheet-11-2020.pdf.
32 This evaluation should include a review of the Defense Production Act (DPA) as a tool for DFC and
potentially other agencies to promote the U.S. industrial base, as was done as part of the response to
COVID-19. See Defense Production Act (DPA), U.S. International Development Finance Corporation
(last accessed Jan. 4, 2021), https://www.dfc.gov/dpa.
33 The Digital Strategy includes several complementary efforts relating to connectivity, cybersecurity,
digital finance, inclusion, and other areas, such as the Digital Connectivity and Cybersecurity
Partnership, Digital Finance, Digital Inclusion, Geospatial Technology and Analytics, Development
Informatics, and Digital Agriculture, among others. Critical components of the Digital Strategy also
include catalytic funding provided to missions (Digital Ecosystem Fund) and conducting Digital
Ecosystem Country Assessments (DECAs). See USAID Digital Strategy 2020-2024, USAID (June
2020), https://www.usaid.gov/usaid-digital-strategy. USAID has also published resources which
are reflective of the Digital Strategy and outline democratic principles in the deployment of those
technologies. See, e.g., Amy Paul, et al., Reflecting the Past, Shaping the Future: Making AI Work
documents/15396/AI-ML-in-Development.pdf; Artificial Intelligence in Global Health: Defining a
in-Global-Health_webFinal_508.pdf.
34 For example, USAID may want to explore the potential for all its programs to include a minimum
threshold of digital programming (e.g., 10% of programmatic efforts include an element of digital
development) as a key element of supporting nations on their journey to self-reliance.
[hereinafter DFC Roadmap for Impact].
36 DFC Roadmap for Impact at 6.
37 DFC Roadmap for Impact at 57; Congressional Budget Justification: Fiscal Year 2021, DFC (last
accessed Nov. 25, 2020), https://www.dfc.gov/sites/default/files/media/documents/FY2021_DFC_
CBJ-Final-04222020.pdf [hereinafter DFC FY2021 Budget].
38 The China Development Bank and China Exim Bank provide concessional loans, including, for
example, a 40-year concessionary loan to Indonesia to fund its U.S. $5.29 B high-speed railway. The
loan provided a 10-year grace period, no guarantees by Indonesia, and local content guarantees. See
China’s Belt and Road Initiative in the Global Trade, Investment, and Finance Landscape, OECD at 18
(2018), https://www.oecd.org/finance/Chinas-Belt-and-Road-Initiative-in-the-global-trade-investment-
and-finance-landscape.pdf.
investment-bank-provide-us84-million-concessional-loan-senegal-support; Japanese Concessional
concessional-oda-loans/; Understanding China’s Belt and Road Infrastructure Projects in Africa,
Brookings Institution (Sept. 2019), https://www.brookings.edu/wp-content/uploads/2019/09/
FP_ 20190930_china_bri_dollar.pdf.
p
555
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
40 DFC has formal relationships with international partners such as Japan and Australia (Japan Bank
for International Cooperation, Nippon Export Investment Insurance, Australia Infrastructure Financing
Facility), the African Development Bank, and the Inter-American Development Bank. See Testimony
by Adam S. Boehler, CEO, U.S. International Development Finance Corporation before the House
Appropriations Subcommittee on State and Foreign Operations, and Related Programs (March 4,
2020), https://www.dfc.gov/testimony-DFC-HAP-03042020.
41 Blended finance, according to the OECD, “is the strategic use of development finance for the
mobilisation of additional finance towards sustainable development in developing countries.” Blended
concessional finance includes the “use of relatively small amounts of concessional donor funds to
mitigate specific investment risks and help rebalance risk-reward profiles of pioneering investments
that are unable to proceed on strictly commercial terms.” See Blended Finance, OECD (last accessed
principles/; Blended Concessional Finance, International Finance Corporation of the World Bank
Group (last accessed Dec. 28, 2020), https://www.ifc.org/wps/wcm/connect/topics_ext_content/ifc_
external_corporate_site/bf; see also Blended Finance, Convergence (last accessed Dec. 28, 2020),
https://www.convergence.finance/blended-finance.
42 This investment fund could be modelled after the $100 million Women’s World Banking Asset
Management Fund, which received $25 million from DFC and $600,000 from USAID. Additionally,
DFC put out a call for proposals for fund managers investing in 5G-related companies operated in
DFC-eligible emerging market countries in order to invest in open and secure ICT. See First-of-Its-
Kind U.S. Government Blended Finance Fund to Empower Women in Developing Markets, DFC (June
finance-fund-empower-women; Information and Communication Technology Call for Proposals, DFC
(March 2020), https://www.dfc.gov/sites/default/files/media/documents/IFD_5G_CFP_032020.pdf.
43 The $82 million request was part of the larger $640.1 million requested to support DDI’s efforts.
See Congressional Budget Justification: Foreign Operations: Appendix 2: FY 2021, U.S. Department
of State at 223 (2020), https://www.state.gov/wp-content/uploads/2021/01/FY21-CBJ-Appendix-2-
FINAL-508-Version.pdf.
44 The Digital Ecosystem Fund “equips the Agency’s Operating Units with catalytic financing to design
and implement activities that foster open, inclusive, and secure digital ecosystems. The DEF supports
two types of interventions: 1) Emergent opportunities to harness or shape the digital ecosystem in any
sector; 2) Strategic initiatives to strengthen or improve the digital ecosystem.” See Digital Ecosystem
Fund: 2020 Activities, USAID (Dec. 22, 2020), https://www.usaid.gov/digital-development/DEF2020.
45 DFC’s FY 2021 budget request sought $700 million in such funds. See DFC FY2021 Budget at 1.
46 The U.S. government announced a first-of-its-kind blended finance fund in June 2019. USAID
provided $600,000 in funding and technical assistance and DFC’s predecessor invested $25 million
to support private capital investments in the $100 million Women’s World Banking Asset Management
Fund. See First-of-Its-Kind U.S. Government Blended Finance Fund to Empower Women in Developing
government-blended-finance-fund-empower-women; Vince Chadwick, USAID, OPIC Team Up on
usaid-opic-team-up-on-women-s-finance-in-preview-of-new-dfi-era-95050.
47 See, e.g., DFC Roadmap for Impact at 57.
48 DFC FY2021 Budget; DFC Roadmap for Impact.
49 DFC Roadmap for Impact at 56.
50 See Chapter 14 of this report for recommendations regarding specific end-use controls on high-end
AI chips.
51 85 Fed. Reg. 43532, Advanced Surveillance Systems and Other Items of Human Rights Concern,
U.S. Department of Commerce: Bureau of Industry and Security (July 17, 2020), https://www.
federalregister.gov/documents/2020/07/17/2020-15416/advanced-surveillance-systems-and-other-
items-of-human-rights-concern; 85 Fed. Reg. 63007, Amendment to Licensing Policy for Items
Controlled for Crime Control Reasons, U.S. Department of Commerce: Bureau of Industry and Security
licensing-policy-for-items-controlled-for-crime-control-reasons.
p
556
BLUEPRINT FOR ACTION: CHAPTER 15
Blueprint for Action: Chapter 15 - Endnotes
52 Guidance on Implementing the UN Guiding Principles for Transactions Linked to Foreign
Government End-Users for Products or Services with Surveillance Capabilities, U.S. Department of
State (last accessed Jan. 4, 2021), https://www.state.gov/wp-content/uploads/2020/10/DRL-Industry-
Guidance-Project-FINAL-1-pager-508-1.pdf.
53 Kara Frederick, Democracy by Design: An Affirmative Response to the Illiberal Use of Technology
reports/democracy-by-design; Dahlia Peterson, Designing Alternatives to China’s Repressive
Surveillance State, Center for Security and Emerging Technology (Oct. 2020), https://cset.georgetown.
edu/wp-content/uploads/CSET-Designing-Alternatives-to-Chinas-Surveillance-State.pdf.
54 GPAI was launched in June 2020 to advance “responsible and human-centric” AI consistent with
human rights, fundamental freedoms, democratic values, innovation, and economic growth. Current
members include Australia, Brazil, Canada, the European Union, France, Germany, India, Italy,
Japan, Mexico, the Netherlands, New Zealand, Poland, Singapore, Slovenia, South Korea, Spain, the
United Kingdom, and the United States, with the OECD and UNESCO as Permanent Observers. See
UNESCO Joins Global Partnership on Artificial Intelligence as Observer, UNESCO (Dec. 10, 2020),
https://en.unesco.org/news/unesco-joins-global-partnership-artificial-intelligence-observer.
55 GPAI’s five working groups (responsible AI, data governance, innovation and commercialization,
the future of work, and pandemic response) are supported by research undertaken by two centres of
expertise: the Paris-based National Institute for Research in Digital Science and Technology (INRIA)
and the Montreal-based International Centre of Expertise in Montreal for the Advancement of Artificial
Intelligence (ICEMAI). See The Global Partnership on Artificial Intelligence Officially Launched,
Montreal International (June 15, 2020), https://www.montrealinternational.com/en/news/the-global-
partnership-on-artificial-intelligence-officially-launched/; Launch of the Global Partnership on Artificial
Intelligence by 15 Founding Members, French Ministry for Europe and Foreign Affairs (June 15, 2020),
global-partnership-on-artificial-intelligence-by-15-founding.
56 The OECD’s work through the Directorate for Science, Technology and Innovation and the AI Policy
Observatory is supported by partnerships with governments and research entities, like the German AI
Observatory’s support of the OECD’s effort on AI’s impact on the labor market. See Work, Innovation,
work-innovation-productivity-skills.
57 Going Digital, OECD (last accessed Feb. 2, 2021), http://www.oecd.org/going-digital/project/.
58 Banff International Research Station (last accessed Jan. 4, 2021), https://www.birs.ca/.
59 AccelNet accelerates network-to-network collaborations by funding the connection (travel,
virtual networking, workshops) between international research networks. NSF only funds the
U.S. portion and expects international partners to fund their part of the collaboration. In addition
to funding connections between existing networks, AccelNet will fund efforts to create and
foster nascent networks. See Accelerating Research Through International Network-to-Network
id=505584&org=OISE&from=home.
60 MULTIPLIER sends subject-matter experts to international areas of interest to assess capabilities
and gather information for potential NSF joint projects. This new program has already been
promising—NSF sent a multidisciplinary team to the Czech Republic and was astounded at their
capabilities. NSF is now exploring bilateral collaboration. See NSF MULTIPLIER: MULTIPlying Impact
Leveraging International Expertise in Research Missions, NSF (last accessed Dec. 28, 2020), https://
www.nsf.gov/od/oise/multiplier.jsp; NSCAI staff discussions with NSF staff (Nov. 14, 2020).
61 For example, the Department of Energy’s national laboratories may be used to sponsor research
with the recommended dedicated funding.
62 The shared research resource can help prevent bottlenecks due to limited compute resources.
This effort may also be, if appropriate, part of an expansion of the National AI Research Resource
delineated in Chapter 9 of this report.
p
557
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
63 NSCAI recommends each member dedicate funding to support research efforts. However, MAIRI
will also serve as a location for research to gather for dialogues, workshops, and mentorships.
Based on similar international research institutes, MAIRI members should consider providing the
equivalent of $100K-$250K per year to cover the travel, accommodations, and per diem of around 80
researchers to MAIRI to facilitate communications and interactions between researchers.
64 In 2017, the government of Alberta (Canada), Canadian Natural Science and Engineering Research
Council, U.S. National Science Foundation, and Mexico’s Consejo Nacional de Cinco y Tecnologia
invested $12.5 million over the next five years. See Research Station Gets $12.5M to Bring Scientists
banff-international-research-station-math-science-funding-1.3977703. For the 2012-2017 period, BIRS
received $10.3 million, of which $3.68 million was from NSF. For the 2006-2011 period, BIRS received
$9.3 million, $3.1 million of which was from NSF. See Organization: Banff International Research
banff-international-research-station/.
65 The National AI Research Institutes provide an important example of the power of leveraging R&D
cooperation across the U.S. interagency. The new NSF and DOE centers for quantum information
science are also a powerful example of the benefits of dedicated resourcing and prioritization across
two U.S. government research entities. See Andrea Peterson, NSF and DOE Support Research
Priorities with Spate of New Center Awards, American Institute of Physics (Sept. 16, 2020), https://
www.aip.org/fyi/2020/nsf-and-doe-support-research-priorities-spate-new-center-awards.
66 For more information, see the Chapter 10 Blueprint for Action.
67 J-1 visas are used by academic employers like universities and research institutions to sponsor
foreign-born academics, interns, trainees, and researchers to work for several months to five years in
the United States. J-1 visa holders are not allowed to renew their visas and must wait one to two years
before they can apply for a different immigration status. O-1 visas are for individuals who can provide
extensive evidence that they have “extraordinary ability in the sciences, arts, education, business, or
athletics.” These visas last for up to three years with indefinite renewals but have been used minimally
to attract experts in S&T due to restrictive policy guidance. For a discussion of immigration and visa
programs to attract scientists and researchers to the United States, see generally Zachary Arnold,
research/immigration-policy-and-the-u-s-ai-sector/.
68 See also Chapter 1 of this report on Malign Information Operations.
69 Secretary Pompeo Approves New Cyberspace Security and Emerging Technologies Bureau, U.S.
Department of State (Jan. 7, 2021), https://2017-2021.state.gov/secretary-pompeo-approves-new-
cyberspace-security-and-emerging-technologies-bureau//index.html.
70 For additional details regarding the function and need for the CSET bureau, see Second Quarter
Recommendations, NSCAI at 88-89 (July 2020), https://www.nscai.gov/previous-reports/.
71 These elements should include key, technology-related functions of the proposed CSET Bureau; the
Bureau of Oceans, Environment and Science (OES); the Office of the Science and Technology Adviser
to the Secretary (STAS); the Coordinator for Cyber Issues (S/CCI); the Bureau for Economic and
Business Affairs (EB); the Bureau for Democracy, Human Rights, and Labor (DRL); and the Center for
Analytics.
72 Key Topics, Office of the Science and Technology Adviser at the U.S. Department of State (last
accessed Dec. 15, 2020), https://www.state.gov/key-topics-office-of-the-science-and-technology-
advisor/.
state.gov/connect/dir/.
74 See, e.g., Managing American Spaces, U.S. Department of State (last accessed Feb. 1, 2021),
https://americanspaces.state.gov/; TechCamp, U.S. Department of State (last accessed Feb. 1, 2021),
https://techcamp.america.gov/; Program Description, World Learning (last accessed Feb. 1, 2021),
https://www.worldlearning.org/program/u-s-speaker-program/.
75 See About GIST, GIST (last accessed Feb. 1, 2021), https://www.gistnetwork.org/about.
p
558
BLUEPRINT FOR ACTION: CHAPTER 15
Chapter 15: A Favorable International Technology Order
Annex: Emerging Technology Coalition
This Annex provides a framework and overarching agenda for global cooperation
on artificial intelligence (AI) and emerging technologies. It includes guidance on concrete,
operational projects; applications; and implementation mechanisms for collaborative AI
work across seven critical areas. Collaborative work in these areas will serve to further AI
consistent with democratic values and strengthen the ties that connect the United States
with its allies and partners. This Annex is intended to provide guidance to the Emerging
Technology Coalition (ETC) and may assist officials in prioritizing bilateral and multilateral
collaborative efforts outside the context, to include engagement with multilateral initiatives
across the AI landscape, as reflected in the Key Multilateral Technology Initiatives Table of
the Chapter 15 Blueprint for Action.
Seven Critical Areas for International Cooperation
STANDARDS
FACILITATE
DEVELOP
PROMOTE & PROTECT
& NORMS
DATA SHARING
AI-RELATED TALENT
INNOVATION
Advance norms
Address legal and
Analyze labor market
Export controls
that uphold
regulatory barriers
challenges; harmonize
Investment screening
democratic values
skills and certification
Supply chain assurance
Explore agreements to
requirements; and
Emerging tech investment
Support secure,
enable data sharing
increase talent
Trade policy
reliable, trusted
and free flow of data
exchanges, joint
Intellectual property
technologies at
training, and workforce
Research and cyber
international technical
development initiatives
protections
standards bodies
JOINT R&D ON
PROMOTE DEMOCRACY,
LAUNCH INTERNATIONAL
AI AND DIGITAL
HUMAN RIGHTS, AND
DIGITAL DEMOCRACY
INFRASTRUCTURE
THE RULE OF LAW
INITIATIVE
Promote collaborative
Pursue joint efforts to
Coordinate international
R&D that advances
counter censorship, malign
foreign assistance, policy
shared interests and
information operations,
and technical guidance,
benefits humanity
human trafficking, and
and development aid
illiberal uses of
and financing
surveillance technologies
p
559
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Critical Area #1 - Developing and Operationalizing Standards and Norms
•
Objectives:
o Advance common, democratic norms and values to govern and guide responsible
artificial intelligence (AI) and the research, development, and application of
emerging technologies globally.1
o Promote international AI norms and standards that uphold democratic values,
building on guiding documents such as the Organization for Economic Co-
Operation and Development (OECD) AI Principles and efforts to operationalize
principles, as reflected in the Commission’s Key Considerations for Responsible
Development & Fielding of Artificial Intelligence.2
o Coordinate positions taken by partner states’ governmental delegations and,
where appropriate, national (non-government) standards institutes accredited to
international technical standards organizations to support development of secure,
reliable, and trusted technologies and to ensure ethical and technical integrity,
endorse standards that comport with democratic values, and maintain the neutrality
of these organizations.3
•
Priority areas for coordination. The ETC should seek to align with allies and partners
and prioritize efforts in development of international technical standards in the following
priority areas:
o Safety and reliability;
o Privacy-enhancing technologies, including privacy-preserving machine learning
(PPML), allied cryptographic code, and other privacy-enhancing technologies;
o Data sharing, labelling, and related documentation for data, models, and systems;
o Assessing system performance and characterizing blind spots per shared values
(including fairness, interpretability, reliability, and secure use of AI technologies as
part of integrated systems);
o Robustness to ensure models are resilient to adversarial examples and model
inversion, while red-teaming with allies on competitors’ attempts to undermine AI-
enabled systems;
o Trust in human-machine teaming and development of common standards and
benchmarks to assess risks in settings of varying complexity and uncertainty;
o Traceability, focused on audit trail requirements per mission needs for high-stakes
AI systems including safety-critical applications; and
o Interoperability, including benchmarks that assess reliability of produced model
explanations.
•
Mechanisms for Coordination.
o In addition to coordinating efforts through the ETC, the U.S. government, led by the
Department of State and in coordination with the National Institute of Standards
and Technology (NIST) and its AI Standards Coordinator, should engage with
democratic nations to align positions on standards that are critical to mutual security
and defense.
p
560
BLUEPRINT FOR ACTION: CHAPTER 15
■ The Department of State, as the Commission has recommended, is in the
process of placing regional technology officers in major foreign technology
hubs.4 This development will facilitate diplomatic efforts toward coordinating
positions with allies and partners.5
■ As the Commission recommended in its Third Quarter Recommendations5
and elsewhere in this Report, NIST and other agencies should consider the
Commission’s Key Considerations for Responsible Development and Fielding
of AI. The Key Considerations include operational guidance on standards
critical to responsible AI and national security, including for technical
standards on testing and evaluation, verification and validation (TEVV).7
o Coordination on technical standards should include the work of international
standards organizations as well as coordinated work on operationalizing AI
norms and principles. The Global Partnership on AI’s (GPAI) Data Governance
Working Group can provide particularly salient best practices for engineering and
implementing data-sharing, pooling, and collecting initiatives.8 The OECD is also a
critical forum for technical standards and guidelines, particularly in data sharing and
responsible AI.9
Critical Area #2 - Joint Research and Development on AI and Digital Infrastructure
•
Objectives:
o Identify areas of shared interest conducive to collaborative R&D—such as privacy-
enhancing technologies, small data approaches to AI, next-generation materials,
prototyping, and high-performance computing (HPC)—for which there are existing
gaps and identify ways to share resources to pursue R&D in those areas.
o Develop mechanisms to facilitate fundamental and applied R&D projects that
involve collaboration among nations, industry partners, and researchers.
■ Projects may also include secure cloud frameworks, sharing best practices
on TEVV, innovative funding models, international test beds to develop
pre-commercial technologies, and leveraging the Commission’s proposed
Multilateral AI Research Institute for coordination (see the Chapter 15 Blueprint
for Action).
o Pursue collaborative, coordinated efforts to develop and deploy AI applications to
benefit humanity at large in areas of global concern such as those embodied in UN
Sustainable Development Goals.
•
Priority areas for collaborative R&D—advancing AI technology.
o Development of privacy-preserving technology, such as homomorphic encryption
and differential privacy techniques,10 to facilitate cross-border AI applications, data
sharing, and cooperative efforts.11
o Continuous development and adaptation of TEVV systems to strengthen the
development of trustworthy, robust AI is critical to advancing the interests of
democratic nations and to understanding how AI systems perform in multi-agent/
adversarial contexts.12 Collaboration in this area will contribute to understanding
differences among allies on policy, metrics, standards, and requirements while
creating stronger connections for all users in a full-cycle approach.
p
561
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
o Development of AI for modeling, simulation, and design to provide researchers with
a larger scope of AI-ready data sets.13
o Development of one- and few-shot learning algorithms14—algorithms that rely
on less data—to facilitate future joint R&D and data sharing and improve context-
specific interoperability.15
o Development of robust allied AI to reduce vulnerabilities of allied AI systems and
training data to adversarial attacks.16
o Achieving context-specific interoperability of AI systems necessary for cross-
border AI applications, with a focus on how systems integrate particular AI/ML
components.17 The potential for AI to increase speed of operations will require
allies and partners to stress-test decision-making procedures and communications
protocols to ensure interoperability. Interoperability of AI systems is already an
issue at the forefront of defense cooperation and will only grow in importance as
technology matures.
o Development of AI to secure and improve resiliency of supply chains to protect AI-
component supply chains while promoting domestic and allied innovation and to
apply AI to improve auditing, mapping, and securing supply chains while ensuring
resilience to shocks. Given the inherently cross-border nature of supply chains and
their critical role in the international economy around AI and advanced technology,
this is a natural area for the United States to work collaboratively with like-minded
nations.
o Additional critical AI research areas including novel machine learning (ML)
directions, complex multi-agent scenarios, advanced scene understanding,
AI system risk assessment, enhanced human-AI interaction and teaming, and
autonomous AI systems.18
•
Priority areas for collaborative R&D—AI to benefit humanity. The potential for AI
to assist the global community in improving the human condition is immense. Priority
areas for international collaboration should include the following:
o Environment and climate. Recognizing the growing view that environmental
degradation and climate change represent imperatives for national and
international security, the international community must work collaboratively to
develop AI-based solutions to address common climate, environmental, and energy
challenges.
■ Collaborative initiatives such as the following serve as models for future
international efforts:
• The Partnership between Cross Section Evaluation Working Group and
OECD’s Nuclear Energy Agency’s (NEA) Working Party on International
Nuclear Data Evaluation Co-operation on International Criticality Safety
Benchmark Evaluation Project19;
• GEOTHERMICA, a collaboration among 12 European countries and the
United States to fund AI-specific research on geothermal R&D20; and
• The International Partnership for Hydrogen & Fuel Cells in the Economy
(IPHE),21 an intergovernmental partnership to facilitate and accelerate
transition to clean and efficient energy with the support of AI and ML
research.
p
562
BLUEPRINT FOR ACTION: CHAPTER 15
o Health, including pandemic detection and response. The COVID-19 crisis has made
clear the need for global collaboration and the potential for AI-enabled solutions.
■
Smart disease monitoring. The Commission has recommended global
cooperation on smart disease monitoring.22 Such a global initiative, for
example, could seek to combine existing data on zoological spills with open
source health-related data to create shared, predictive, global disease
monitoring models (see Chapter 16 of this report and its associated Blueprint
for Action).23
■
Pandemic preparedness, vaccine development, and syndromic surveillance.
Efforts in this space24 include:
• Development and coordination on international norms and standards to
govern use and sharing of international health data, protecting privacy
while ensuring timely accessibility of data;
• Development of privacy standards for genomic data sets;
• Increased international cooperation in the COVID-19 High Performance
Computing Consortium (potentially through GPAI); and
• Facilitation of international cooperation with DARPA’s work on creating the
infrastructure and protocols for data sharing and collaboration at the point
of experimentation for drug discovery.
■
Initiatives to enable long-term quality of life. Collaboration with allies and
partners can facilitate the Commission-recommended focus on harnessing AI
to help the elderly live independently longer, assist in managing health and
daily tasks, and improve the quality of life, particularly through the application
of AI to biomedicine.25
26
■
The National Nanotechnology Initiative’s U.S.-EU Communities of Research,
along with various national-level efforts by partner nations,27 should serve as
models for larger-scale international collaboration.
o Food security. The United States emphasizes agriculture-led growth, resilience,
nutrition, and water security, sanitation, and hygiene in its foreign assistance
programs. Enhancing the security of water and food of partner nations is needed to
disrupt the vicious cycle of poverty, hunger, and conflict.28
■ Agricultural sectors are increasing the use of data-driven technologies such as
robotics, satellites, GPS, and drones. Significant data sets are being generated
about crop growth, soil characteristics, and weather conditions.29
■ AI and ML-based algorithms can amplify the data sets and hardware to improve
real-time monitoring and analysis of agricultural and distribution processes.
This can improve efforts to assess needs, enhance productivity and security,
and build local capacities and productivity while minimizing environmental
impact.
o Disaster relief. AI-enabled technologies are being used to address a range of
disaster scenarios, and further work on an international basis should be explored.
The World Economic Forum is among those groups calling for greater international
collaboration in order to realize the benefits of AI to specifically include the area of
disaster relief.30
p
563
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
■ Climate- and weather-related disasters like hurricanes, wildfires, and flooding
are on the rise, and AI is already being applied to mitigate the effects by
locating survivors using unmanned aerial vehicles, removing debris afer a
disaster, deploying robots to communicate with victims, employing edge
technology to obtain the most up-to-date data, running predictive simulations,
and leveraging social media reporting.
■ The DoD is partnering with DoE and Microsof to develop “deep-learning
artificial intelligence algorithms to provide near-real-time data to improve
the decision-making of first responders engaged in natural disasters and
humanitarian assistance efforts,”31 and countries like Singapore are working
with the Joint Artificial Intelligence Center (JAIC) on this particular National
Mission Initiative.32
■ The United States can work with its allies and partners to effectively predict,
model, prepare for, and respond to disasters, as the Commission recommends
in Chapter 11 of this report.33
o Civilian space cooperation. The United States and other space agencies employ AI
to tackle a range of space missions—including for visualization of space objects and
situational awareness, tracking space debris for satellite collision avoidance, roving
the lunar surface, deep-space exploration with autonomous systems, and detection
of asteroids that could threaten Earth.34
■ For safe satellite navigation around space debris, the European Space Agency
(ESA) and the UK Space Agency both have AI initiatives underway, which
suggests potential opportunities for closer U.S. collaboration.35 ESA has
already established a partnership with Stanford.36
■ India is also building its space program and has deployed an AI-powered Moon
rover.37
■ Russia and China appear to be developing technological solutions to the
space collision problem, which could present an area for mutually beneficial
cooperation. AI-enabled robotic assistants are also being developed for the
International Space Station.38
•
Methods to implement collaborative R&D. The ETC should also explore vehicles
to enable R&D collaboration among government partners and non-governmental
organizations.
o Collaboration must include not only government-to-government efforts, but also
methods to partner with researchers at academic research centers and in the private
sector.
o Existing science and technology (S&T) agreements between governments may
provide the legal foundation for cooperation, but details will depend on the
arrangement at issue.
o The ETC should prioritize approaches that would facilitate the pooling of resources,
reduce redundancies, and support development and socialization of best practices.
o In addition, the ETC should examine challenges to cross-border, collaborative
R&D—such as those around data privacy and data sharing between the United
States and European Union—and explore solutions to overcome those challenges.
p
564
BLUEPRINT FOR ACTION: CHAPTER 15
•
Potential methods to implement and further collaborative R&D include:
o Establishment of the Multilateral AI Research Institute (MAIRI). Proposed by the
Commission in this report, MAIRI will serve as a center for multilateral research
to coordinate joint efforts to develop technologies and align norms that advance
responsible, human-centric, and privacy-preserving AI/ML that better societies.
o Prioritization of R&D work of the Global Partnership on AI (GPAI). The ETC should
leverage existing frameworks wherever possible, and GPAI, supported currently
by Canada’s International Center of Expertise in Montréal for the Advancement of
Artificial Intelligence (ICEMAI) and France’s National Institute for Research in Digital
Science and Technology (INRIA),39 is among the most promising multilateral, multi-
stakeholder initiatives to pursue collaborative R&D and advance AI technology for
common causes. The Commission has proposed a greater role for U.S. researchers
through a U.S.-based Center of Expertise, leveraging the NSF National AI Research
Institutes.40
o Creation of a joint emerging-tech investment consortium. Modeled on In-Q-Tel,
the consortium would spur investment by the United States and foreign partners
in early-stage companies to further development of AI technology that advances
and/or protects democratic values. The effort would benefit the United States and
its allies and partners through a cross-border platform to engage with startups and
entrepreneurs in the AI and emerging-tech space.
■ Within the U.S. government, this effort should draw on State Department’s
Regional Technology Officers, the Foreign Commercial Service, and
USAID missions to identify R&D and prototypes to advance U.S. diplomatic,
development, and commercial interests.
o Launching multilateral innovation prize competitions. Modeled on Defense
Advanced Research Projects Agency (DARPA) Challenges and XPRIZE Foundation
competitions, international innovation prize competitions sponsored by two or
more governments would incentivize R&D in fundamental AI or around specific
applications necessary for national security and help to pool resources and talent
with allies and industry.41
o Fostering allied research at U.S. national labs. The ETC should consider
recommendations for increasing research by allies (potentially a subset of the ETC
membership) at U.S. national labs on sensitive topics. Although there are limitations
on U.S. national labs to allow foreign researchers, domestically housed research
efforts would limit concerns around cross-border data-sharing and cybersecurity
and could prove fruitful in R&D necessary for defense and security applications.
o Development of an R&D matching platform and a global horizon-scanning
capability. The platform would connect researchers and their projects with funders
and partners (governments, philanthropists, venture capitalists, companies,
research institutions), providing the U.S. Government with increased visibility into
research trends. A horizon-scanning capability of global R&D would complement
these efforts and draw on open-source data to give policymakers greater
understanding of relevant discoveries and key trends in the field.42
o Development of an international test bed for TEVV. An international test bed
for TEVV could be modeled on the National AI Research Institutes43 or the
Commission-recommended creation of a NIST-sponsored third-party testing
center,44 but with a cross-border focus, as well as on the AI4EU project.45
p
565
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
o Improved collaboration between centers of excellence, research institutes, and
industry consortia through additional coordination by partner governments.
This concept would leverage existing and soon-to-be-established centers like
the MAIRI, European Union Centres of Excellence, European AI-related Digital
Innovation Hubs, the U.S. National AI Research Institutes program, the General
Services Administration’s AI Center of Excellence (in partnership with the
Departments of Agriculture and Health and Human Services as well as the JAIC),
the Alan Turing Institute in the U.K., and in Canada the Montreal Institute for
Learning Algorithms (Mila),46 the Alberta Machine Intelligence Institute (Amii),47 and
the Vector Institute for Artificial Intelligence.48 On the U.S. side, this could involve
building on industry and academic efforts like the Stanford Institute for Human-
Centered Artificial Intelligence.49
o Fostering of binational R&D foundations. ETC members may consider developing
targeted, binational R&D efforts modeled on the unique binational foundations that
facilitate U.S.-Israel and U.S.-India R&D on cutting edge issues.50 These can serve
as models for other allies and partners to convene international researchers.
Critical Area #3 - Promoting Democracy, Human Rights, and the Rule of Law
•
Objectives:
o Collaborative, coordinated efforts to counter anti-democratic uses of AI and
emerging technologies through coordinated policy, regulatory alignment (such as
end-user export restrictions), and technology deployment.
o Potential priorities include countering censorship, countering malign information
operations, and promoting democratic models of surveillance technology, although
the ETC should explore a range of potential applications.
o Furthering these normative priorities will build on implementation methods
addressed in other Critical Areas, especially #2 (joint R&D), #5 (protecting and
promoting innovation), and #7 (the International Digital Democracy Initiative).
•
Countering censorship and authoritarian uses of technology
o The ETC should explore efforts to use AI and associated technologies to further
internet freedom and counter censorship across the world. This work should be
designed to implement the principles adopted in November 2020 by the Freedom
Online Coalition, a partnership of 32 governments aligned around promoting
human rights and individual freedom.51
o The United States should leverage the Open Technology Fund, created by the
FY 2021 NDAA, to support this effort, as well as related efforts by the Bureau of
Democracy, Human Rights, and Labor (DRL) at the Department of State.52
o The ETC should coordinate efforts in this space with the Council of Europe’s Ad Hoc
Committee on AI, established in November 2019 to focus on development, design,
and application of AI in areas of human rights, democracy, and the rule of law.53
o To promote private-sector conduct that comports with shared democratic values,
the ETC should develop a proposal for end-user controls that would disincentivize
private companies from exporting AI and associated technologies that may be used
to suppress and violate human and civil rights.54
p
566
BLUEPRINT FOR ACTION: CHAPTER 15
•
Countering Malign Information Operations
o Malign information operations present a growing international challenge that is
compounded by the use of AI/ML technologies.
o This ETC should examine coordinated efforts (outside of the intelligence space)
to counter disinformation and other information operations. Joint efforts include
detecting, moderating, identifying, and classifying malign information, developing
standards and best practices, and training experts.
o The Commission recommends creation of an International Task Force to Counter
and Compete Against Disinformation.55 An International Task Force to Counter and
Compete Against Disinformation (ITF-CCAD) could be established as a joint project
between the United States and multiple countries, as well as the EU and NATO, to
further joint efforts to enable content moderation and detection of disinformation,
develop standards for identifying and classifying misinformation and disinformation
(to include deepfake detection), and share best practices and lessons learned
with allies. The private sector, academia, and civil society organizations would be
important partners in this effort.
o The ITF-CCAD should draw best practices from, and should work in coordination
with, the Global Internet Forum to Combat Terrorism,56 along with efforts of the
Department of State’s Global Engagement Center’s (GEC) Technology Engagement
Team (TET); the Federal Bureau of Investigation’s Foreign Influence Task Force
(FITF); the European External Action Services’ Strategic Communication Task Force;
the EU “Team Europe” initiative; and the NATO/StratCom Center of Excellence. IFT-
CCAD should additionally prioritize stress-testing rapid-response mechanisms and
look to fund open-source research.
o It should explore generating best practices for non-tech solutions, such as media
literacy, free press,57 and civic engagement initiatives, drawing on notable work
by the Center for Strategic and International Studies’ Defending Democratic
Institutions project and the German Marshall Fund’s Alliance for Securing
Democracy.
•
Surveillance technology that comports with democratic values.
o The ETC should dedicate a multifaceted effort to promoting surveillance
technology that supports democratic values.58 In particular, the effort should focus
on (a) promoting technology that delivers a degree of protection for individual
privacy and for civil rights and civil liberties and limits the use of data collected
or combined in ways that enable re-identification, and (b) countering the global
deployment of surveillance technology used to undermine democratic values and
individual rights.
o Doing so will require coordinated R&D, messaging, and development assistance
strategies to support democratic alternatives to technology manufactured in
China.59
o Fostering the R&D necessary to provide alternatives will require public-private
coordination or partnerships at an international level (see Critical Area #2 for
potential mechanisms).
■ Potential stakeholders for such a project include NSF, the National AI Research
Institutes, DARPA, NIST, various EU Centres of Excellence, research institutions
(such as the Johns Hopkins University Applied Physics Laboratory, the
p
567
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Massachusetts Institute of Technology Computer Science & AI Laboratory, and
the Stanford Institute for Human-Centered AI), GPAI, and non-governmental
organizations such as OpenMined.
Critical Area #4 - Exploring Ways to Facilitate Data Sharing
•
Objectives:
o Address legal and regulatory barriers to international collaborative work; explore
bilateral and multilateral, general and specific approaches to enable data sharing,
pooling, and storing consistent with privacy, security, and other fundamental values,
including the viability of a Data Free Flow with Trust Agreement.
•
Methods to implement coordinated approaches to data sharing.
o Development of shared data environments. Development of pooled data storage
centers, computational environments, and cloud and edge computing facilities
to pool data from different sources for free use by credentialed researchers. An
approach like this would prove particularly beneficial to improve data sharing
among members to the Five Eyes alliance.
o Agreement on foundational data documentation, labelling, archiving, and data
organization frameworks at international organizations. Data agreements among
members of alliances (such as NATO) or other international organizations would
facilitate support to collaborative R&D endeavors; for example, ongoing efforts at
the Organisation of Economic Co-operation and Development (OECD) AI Policy
Observatory and Global Partnership on Artificial Intelligence (GPAI).
o Agreements to share specific data sets with specified foreign partners. Narrower
in scope than the above two approaches, an agreement of this kind would allow
researchers from different countries to access the same data sets for their respective
projects. For example, in the context of COVID-19 and health care,60 countries
would need to address data labelling, data storage, data anonymization, data
security, and other issues on a joint basis or through a pilot project.
61
■ This type of effort could also include joint projects with allies to anonymize
high-impact data sets for specific research or initiatives, such as National
Institutes of Health data sets and data sets maintained, for various purposes, by
DOE, the U.S. Agency for International Development (USAID), the U.S. Food
and Drug Administration, DARPA, IARPA, and the Department of State’s Center
for Analytics.
■ Diplomatic effort is needed to resolve divergent views over what constitutes
anonymized data, consent, and matters of public interest.
o Ad hoc data sharing arrangements on bilateral or multilateral bases. The ETC
should explore the willingness of strategic allies and partners to engage in targeted,
non-treaty data-sharing arrangements. “Innovation sandbox” arrangements may
be designed to facilitate specific challenges across all domains—security, health,
disinformation, environmental resilience, and so on.
o A multilateral data-sharing agreement founded on trust. The ETC should lead
an effort to create a formal, potentially treaty-based approach to data sharing,
pooling, and storing with like-minded governments modeled on the data free flow
with trust (DFFT) concept introduced by Japan at the June 2019 G20 Summit.
p
568
BLUEPRINT FOR ACTION: CHAPTER 15
DFFT would permit the free flow of data between authorized parties upon meeting
specific standards, including intellectual property (IP), privacy, and cybersecurity
protections.62 The European Commission endorsed the DFFT concept in December
2020.63
■ A general DFFT would require significant consideration of data protection,
IP protection, privacy shield, and trade issues, both for the United States
domestically and for foreign partners.
■ A specific DFFT, on the other hand, focused on the free flow of data for
particular purposes—such as facilitating pandemic response efforts—would
have a greater chance of success and could be a model for targeted data-
sharing arrangements in other areas of shared interest.
o Development of a secure AI research resource infrastructure. A secure, cloud-based
infrastructure would provide researchers from partnered and allied countries access
to compute resources, diverse data sets, and controlled environments to enable
testing, for example of privacy-preserving ML techniques. Participating like-minded
governments would agree to and comply with common technical standards and
norms64 and risk-based frameworks that ensure privacy, security, reliability, respect
for the rule of law, and other appropriate parameters.
■ Such an infrastructure could be developed bilaterally or multilaterally and
could be a priority effort of the Commission’s proposed MAIRI. Research and
academic institutions could support the MAIRI effort with appropriate technical
and implementation assistance, while GPAI’s Data Governance Working Group
could support the development and utilization of engineering best practices.
Critical Area #5 - Promoting and Protecting Innovation
•
Objectives:
o Develop an allied strategy to align and develop regulatory and legal regimes
in areas critical to fostering domestic and international innovation. These areas
include export controls, investment screening, supply chain assurance, emerging
technology investment, trade policy, IP, technology transfer, and research
protection.
o Achieving such a strategy will require an integrated approach among allies and
partners, leveraging our full technology toolkit, upgrading capabilities and, where
necessary, developing new ones to counter threats. These efforts will require a
coordinated strategic coordination plan to raise allied public awareness on issues
such as technology-transfer risks.
•
Export Controls and Investment Screening
o The ETC should explore coordinated approaches to export controls and investment
screening. Cooperation in these areas is critical to ensure that like-minded nations
have the authority to unilaterally institute export controls and block predatory
investments that present risks to national and international security.
o The Commission recommends in Chapter 14 of this report that the United States
engage with allies and partners on legal reforms to (a) implement a coordinated
approach to AI-related export controls and (b) enhance investment screening
procedures and enforcement.65
p
569
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
o The Commission has also recommended as part of Chapter 15’s Blueprint for
Action that the United States should engage with allies and partners to align policy
guidance on exports as part of the International Digital Democracy Initiative (IDDI) to
promote technologies that comport with shared values and support free and open
societies.66
o As detailed in Chapter 14 of this report, export control priorities should include
targeted, high-end semiconductor manufacturing equipment (SME) components
needed to produce chips at the 16nm node and below. Additionally, states should
explore implementing targeted end-use and end-user controls on specific high-
end, AI-specialized chips to prevent their use in human rights violations.67
o Consideration should be given to appropriate economic incentives to support
alignment on export control and investment screening.
o The ETC should also pursue robust collaboration on foreign S&T and investment
flow monitoring—to include open-source intelligence—to utilize early warning
indicators related to strategic acquisition risks. Further, ETC partners must share
best practices to monitor smaller transactions that attempt to skirt existing controls.
•
Supply Chain Assurance
o Leadership of the United States and its allies and partners in emerging technologies
is dependent on components sourced from strategic competitors or regions with
significant geopolitical risk.
o The semiconductor manufacturing industry is a prime example of an industry that
is critical to U.S. and allied security, but which is heavily concentrated in specific
geographic regions and is therefore susceptible to supply chain shocks, particularly
in the event of a crisis.68
o The ETC should conduct a supply chain assessment and make recommendations
on integrated, multilateral approaches to coordinating critical technology
components to enhance international security while reducing collective
dependence on strategic competitors.69
o The ETC should also develop a strategic plan to fund key choke point technologies
and next-generation materials, approaches, and prototyping capabilities at
discovery, manufacturing, and applied scales.70
•
Emerging Technology Investments
o Likewise, investments in emerging technologies require coordinated action. 5G
presents a test case for the challenges of international and multilateral coordination.
The United States and partners have cooperated on developing alternatives to
Chinese 5G infrastructure multilaterally and bilaterally.
o The Commission offered recommendations regarding steps to promote domestic
development of 5G technology in its First Quarter Recommendations and urged
the United States to continue to work closely with key allies and partners on both
constructive 5G technical solutions, and to ensure that global 5G networks are
safe and secure.71 Chapter 16 of this report details steps to promote domestic
development of biotechnology, 5G, quantum computing, autonomy and robotics,
advanced manufacturing, and energy systems, while Chapter 13 of this report
details steps to cultivate domestic innovation in microelectronics research and
manufacturing.72 The United States may engage key allies and partners on these
technologies.
p
570
BLUEPRINT FOR ACTION: CHAPTER 15
o The ETC can serve as a forum to explore these issues in a coordinated manner.
•
Trade Policy
o Trade policy is a key lever for the United States and foreign governments to promote
an innovation environment. The ETC should consider coordinated approaches to
trade policy to further innovation and strengthen national and international security.
•
Intellectual Property
o IP rights and regimes are critical to innovation in AI and emerging technologies.
The ETC should explore coordinated approaches to IP that could inform a mutual
agenda with the World Intellectual Property Organization’s (WIPO) Conversation on
AI and Intellectual Property, IP5,73 and forums with broader mandates.
o Coordination on assistance to nations in developing strong and aligned IP regimes.
The ETC can assist the United States and partners in prioritizing assistance to
nations in improving their IP regimes to help facilitate innovation while deterring
IP thef. A more focused approach, through IP5 and WIPO, may prove more
impactful in scope and could help to harmonize efforts to shore up IP with respect to
identifiable international challenges.
■ The United States should engage with key allies and partners to align
on critical aspects of IP, including patent eligibility for AI and associated
technologies, countering China’s narrative on winning the innovation
competition, IP contractual ecosystem impediments to international
collaboration, IP protections for data, and the over-declaration of “standard
essential” patents and other efforts to efforts by countries to exploit standards-
setting and licensing processes.74
■ These are among a set of 10 critical IP considerations that the Commission
proposes to guide U.S. efforts to reform IP policies and establish new IP regimes
for AI and critical emerging technologies in order to protect and promote
national security, innovation, and technology competitiveness.75
o Coordinated efforts to stop IP thef and counter cyber espionage. IP thef remains
a global concern. With a goal of protecting the economic viability of AI innovation
and emerging technologies, the ETC should identify methods to strengthen the
international framework for addressing the export of counterfeit goods, thef of IP
technology, forced technology transfers of foreign innovation, and cyber espionage.
•
Research and Cyber Protections
o Promoting multilateral responses to research integrity and security. As the
Commission has proposed, the United States should coordinate action with allies
and partners in developing multilateral responses to challenges to research
integrity and security posed by PLA-affiliated individuals and entities and to
promote a commitment to open fundamental research.76
■ A public-private research security clearinghouse that enables sharing of open-
source information, data-driven assessments, decision-support resources, and
education and training resources could strengthen this effort.77
o Promoting multilateral efforts to mitigate proliferating cyber vulnerabilities and
develop AI-enabled defenses against cyber attacks. As the Commission has
proposed, the United States must prepare for AI-enabled cyber conflict. The United
States should explore coordinating and joint efforts with key allies and partners.78
p
571
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
Critical Area #6 - Developing AI-Related Talent
•
Objectives:
o Cooperative efforts to enable government, military, academic, and private-
sector talent exchanges and address challenges posed by immigration and visa
restrictions; development of joint AI and digital training and workforce-development
programs.
•
Methods for furthering talent development globally. The ETC should explore
methods for achieving objectives, including the following:
o Creating new models for international talent exchanges. International talent
exchanges are powerful tools to further AI alignment, cross-pollinate ideas, and
build AI-related skills and capabilities. In developing new approaches to talent
exchanges, the ETC should consider:
■ Military officer exchanges to improve AI deployment and interoperability,
including among NATO, JAIC, DoD, and foreign defense ministries and
militaries;
■ Analogous training and exchanges needed for U.S. and allied diplomats and
development experts;
■ Government-to-government exchanges of AI experts to assist in building
tech and ethical expertise; exchanges to benefit industry-led multilateral and
multi-stakeholder efforts like SDOs, GPAI, OECD, and influence paths taken by
partners;
■ Talent exchanges and secondments in industry and academia (both
international industry-industry or academia-academia talent exchanges, as
well as government-industry/academia); and
■ Leveraging research centers such as the proposed MAIRI to enable cross-
border collaboration and talent exchanges.
o Coordinating AI training development programs and sharing of best practices for
government training and broader AI education programs (including in secondary
schools and universities to include computer science teaching and curriculum
development).
■ The ETC should explore methods for non-EU partner nations to coordinate on
the “Artificial Intelligence and Analytics” in the EU’s Digital Education Plan.
Critical Area #7 - International Digital Democracy Initiative
• The Chapter 15 Blueprint for Action details the Commission’s recommendations for
coordinating foreign assistance, investment, and financing through the International
Digital Democracy Initiative.
p
572
BLUEPRINT FOR ACTION: CHAPTER 15
Chapter 15: A Favorable International Technology Order
Annex: Emerging Technology Coalition - Endnotes
nscai.gov/previous-reports/.
2 See the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details on the Commission’s
recommendations for upholding democratic values, see the section on “Aligning Systems and Uses
with American Values and the Rule of Law” in Key Considerations for Responsible Development &
Fielding of Artificial Intelligence: Extended Version, NSCAI (2021) (on file with the Commission).
3 Key technical standards organizations include the International Organization for Standardization
(ISO), International Electrotechnical Commission (IEC), Institute of Electrical and Electronic Engineers
(IEEE), the UN International Telecommunication Union’s Telecommunication Standardization Sector
(ITU-T), and the Third Generation Partnership Project (3GPP). See Interim Report and Third Quarter
Recommendations, NSCAI at 205 (Oct. 2020), https://www.nscai.gov/previous-reports/.
reports/.
nscai.gov/previous-reports/. These recommendations have been reprised and built upon in the
Chapter 15 Blueprint for Action.
6 Id.
7 The Commission recommends that NIST provide a set of standards, performance metrics, and tools
for qualified confidence in AI models, data, and training environments and predicted outcomes. The
Blueprint for Action also recommends that NIST provide guidance as the science on testing across
responsible AI attributes evolves. See the Chapter 7 Blueprint for Action.
8 Jenni Tennison, An Introduction to the Global Partnership on AI’s Work on Data Governance,
OECD AI Policy Observatory (Aug. 21, 2020), https://oecd.ai/wonk/an-introduction-to-the-global-
partnership-on-ais-work-on-data-governance.
9 The OECD has led the international community with its work around AI norms and policy
development. The May 2019 Principles on Artificial Intelligence was the first multilateral set of
principles adopted by governments. Launched in February 2020, the OECD AI Policy Observatory
facilitates dialogue between its global multi-stakeholders, provides evidence-based analysis on
20+ policy areas, promotes the adoption of the AI Principles, and bolsters the advancement and
monitoring of trustworthy AI systems that benefit society. The Network of Experts on AI (ONE AI)
is an informal advisory group of multidisciplinary and multi-stakeholder experts from more than
30 countries that provides policy, technical, and business expert input to inform OECD analysis
and recommendations. The OECD has also developed Guidelines on the Protection of Privacy and
Transborder Flows of Personal Data and is working toward principles for trusted government access
ieconomy/privacy-guidelines.htm.
10 See the Technical Glossary of AI Terms Appendix of this report for definition of homomorphic
encryption and differential privacy techniques. Collaborative research in this area could draw
from promising R&D use cases, including the DARPA Brandeis program and the IARPA HECTOR
brandeis; Homomorphic Encryption Computing Techniques with Overhead Reduction (HECTOR),
IARPA (last accessed Sept. 18, 2020), https://www.iarpa.gov/index.php/research-programs/hector;
see the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details on the Commission’s
recommendation for future R&D needed to advance capabilities for preserving and ensuring
American values and the rule of law, see the section on “Aligning Systems and Uses with American
Values and the Rule of Law” in Key Considerations for Responsible Development & Fielding of
Artificial Intelligence: Extended Version, NSCAI (2021) (on file with the Commission).
p
573
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
reports/.
12 See the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details on the Commission’s
recommendation for future R&D needed to advance capabilities for TEVV, see the section on “System
Performance” in Key Considerations for Responsible Development & Fielding of Artificial Intelligence:
Extended Version, NSCAI (2021) (on file with the Commission). TEVV R&D includes complex system
testing to improve understanding of and confidence in emergent performance of composed AI
systems and improve methods to understand, predict, and control systems-of-systems to avoid
negative outcomes resulting from system interaction. In addition, R&D in a multi-agent scenario will
advance the understanding of interacting AI systems, including the application of game theory to
varied and complex scenarios and interactions between cohorts composed of a mixture of humans
www.nscai.gov/previous-reports/.
13 The Commission has previously recommended that the United States devote greater resources to AI
modeling, simulation, and design. See First Quarter Recommendations, NSCAI at 6-13 (March 2020),
https://www.nscai.gov/previous-reports/.
14 See the Technical Glossary of AI Terms Appendix of this report for definition of one-shot (or few-
shot) learning.
reports/.
16 See the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details on the Commission’s
recommendation for future R&D needed to advance capabilities for AI security and robustness, see
the section on “Engineering Practices” in Key Considerations for Responsible Development & Fielding
of Artificial Intelligence: Extended Version, NSCAI (2021) (on file with the Commission). See also First
Quarter Recommendations, NSCAI at 11 (March 2020), https://www.nscai.gov/previous-reports/.
17 See the Appendix of this report containing the abridged version of NSCAI’s Key Considerations
for Responsible Development & Fielding of AI. For additional details on the Commission’s
recommendations to test machine-machine/multi-agent interaction and for international collaboration
and cooperation needed to align on how to test and verify AI system reliability and performance along
shared values , see the section on “System Performance” in Key Considerations for Responsible
Development & Fielding of Artificial Intelligence: Extended Version, NSCAI (2021) (on file with the
Commission).
18 See Chapter 11 of this report and its associated Blueprint for Action for detail on potential priority
areas for AI research.
www.oecd-nea.org/science/wpncs/; International Co-operation in Nuclear Data Evaluation: An
Extended Summary of the Collaborative International Evaluated Library Organisation (CIELO) Pilot
wpec/documents/volume40.pdf.
20 GEOTHERMICA combines financial resources and expertise on geothermal energy research
and innovation from 16 countries and their regions. It “launches joint projects that demonstrate
and validate novel concepts of geothermal energy deployment within the energy system, and that
identify paths to commercial large-scale implementation.” One of the three focus areas includes
“smart integration into the energy system and operations.” Some of the projects have big-data and
smart-system aspects, such as the French National Project through the GEOTHERMICA HEATSTORE
project. GEOTHERMICA partners, like the U.S. Lawrence Livermore National Laboratory, have
www.geothermica.eu/about-geothermica/; French National Project, HEATSTORE (last accessed Sept.
18, 2020), https://www.heatstore.eu/national-project-france.html; American Partners, GEOTHERMICA
(last accessed Sept. 18, 2020), http://www.geothermica.eu/matchmaking/united-states/.
p
574
BLUEPRINT FOR ACTION: CHAPTER 15
Chapter 15: A Favorable International Technology Order
Annex: Emerging Technology Coalition - Endnotes
21 Members of the partnership include the United States as well as Australia, China, Germany, Japan,
Russia, Austria, Costa Rica, Iceland, ROK, South Africa, Brazil, India, the Netherlands, Canada,
France, Italy, and Norway. See International Partnership for Hydrogen and Fuel Cells in the Economy,
U.S. Department of Energy (last accessed Sept. 18, 2020), https://www.energy.gov/eere/fuelcells/
international-partnership-hydrogen-and-fuel-cells-economy; International Partnership for Hydrogen
and Fuel Cells in the Economy, IPHE (last accessed Sept. 18, 2020), https://www.iphe.net/.
nscai.gov/previous-reports/.
23 Chapter 16 and its associated Blueprint for Action recommends that the United States pursue
global cooperation on smart disease monitoring.
24 Jason Matheny, et al., The Role of AI Technology in Pandemic Response and Preparedness:
papers/covid-19-white-papers/.
25 See Chapter 11 of this report and its associated Blueprint for Action for its recommendation for the
United States to take on some of humanity’s biggest challenges.
26 NanoEHS CORs, US-EU Nanotechnology Communities of Research (CORs) (last accessed Sept.
18, 2020), https://us-eu.org/communities-of-research/.
27 See, e.g., Jeff Mason, et al., An Overview of Clinical Applications of Artificial Intelligence, Canadian
Agency for Drugs and Technologies in Health (Sept. 2018), https://www.cadth.ca/sites/default/files/
pdf/eh0070_overview_clinical_applications_of_AI.pdf.
28 Feed The Future (last accessed Feb. 3, 2021), https://www.feedthefuture.gov/.
29 Automation and Artificial Intelligence in Agriculture: The Future of Maintaining Food Security
and Sustainable Intensification, Frontiers (last accessed Feb. 3, 2021), https://www.frontiersin.org/
research-topics/15206/automation-and-artificial-intelligence-in-agriculture-the-future-of-maintaining-
food-security-and-su.
30 The World Economic Forum has noted 160 million people a year are at risk from natural disasters
and sees great benefit in AI from “reducing the time to assess damage to monitoring social media
to more quickly and effectively deliver aid” while “sharpen[ing] the decisions of relief workers on the
front lines.” Ashley van Heteren, et al., Natural Disasters are Increasing in Frequency and Ferocity.
weforum.org/agenda/2020/01/natural-disasters-resilience-relief-artificial-intelligence-ai-mckinsey/.
31 David Vergun, DOD Partners With Agencies to Use AI for Disaster Relief, Humanitarian Relief, DOD
News (Aug. 20, 2020), https://www.defense.gov/Explore/News/Article/Article/2319945/dod-partners-
with-agencies-to-use-ai-for-disaster-humanitarian-relief/.
32 Prashanth Parameswaran, What’s in the New US-Singapore Artificial Intelligence Defense
singapore-artificial-intelligence-defense-partnership/.
33 See the discussion in Chapter 11 and its associated Blueprint for Action on using AI to tackle some
of humanity’s biggest challenges.
deep-asteroid/.
35 AI Challenged to Stave off Collisions in Space, European Space Agency (Oct. 9, 2019), https://www.
esa.int/Enabling_Support/Space_Engineering_Technology/AI_challenged_to_stave_off_collisions_
in_space; Angelica Mari, UK Government Seeks Innovations to Tackle Space Debris, Computer
Weekly (May 28, 2020), https://www.computerweekly.com/news/252483762/UK-government-seeks-
innovations-to-tackle-space-debris.
p
575
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
36 Andrew Myers, Stanford Develops an AI Navigation System for a Future Satellite ‘Tow Truck,’
Stanford News (Feb. 1, 2019), https://news.stanford.edu/2019/02/01/stanford-spurs-ai-navigation-
space-rendezvous-software/.
37 Leslie D’Monte, Chandrayaan-2 Pragyan Shows How AI is Helping Space Exploration, Mint (Sept.
helping-space-exploration-1567764065716.html.
38 Mike Wall, New, Emotionally Intelligent Robot CIMON 2 Heads to Space Station, Space.com (Dec. 5,
2019), https://www.space.com/cimon-2-artificial-intelligence-robot-space-station.html.
39 The EU intends to establish Centres of Excellence and Digital Innovation Hubs focused on AI.
ICEMAI works with the Government of Canada’s Advisory Council on Artificial Intelligence, Forum
IA Quebec, and the International Observatory on the Societal Impacts of Artificial Intelligence and
Digital Technologies and is supported by the governments of Canada and Quebec with up to $15
million in funding over five years. INRIA was launched in February 2020 and has a contract with the
Government of France to focus on “speeding up development of France’s scientific and technological
leadership, as part of a Europe-wide approach,” including prioritizing AI and other digital
technologies to meet societal challenges, constructing European research and innovation spaces,
strengthening the tech industrial base, reinforcing public policies, and developing leading research
universities. See Communication Artificial Intelligence for Europe, European Commission (April 25,
europe; The Global Partnership on Artificial Intelligence Officially Launched, Montreal International
(June 15, 2020), https://www.montrealinternational.com/en/news/the-global-partnership-on-artificial-
intelligence-officially-launched/; INRIA: For Scientific, Technological and Industrial Leadership in
scientific-technological-and-industrial-leadership-in-digital-technology.
40 See Implementation Plan to Chapter 15: A Favorable International Digital Order.
artificial-intelligence.
42 Melissa Flagg & Paul Harris, System Re-engineering: A New Policy Framework for the American
cset.georgetown.edu/research/system-re-engineering/.
43 The National AI Research Institutes is a joint government effort among the National Science
Foundation (NSF), U.S. Department of Agriculture (USDA) National Institute of Food and Agriculture
(NIFA), U.S. Department of Homeland Security (DHS) Science & Technology Directorate (S&T), and
the U.S. Department of Transportation (DOT) Federal Highway Administration (FHWA). See National
Artificial Intelligence (AI) Research Institutes, NSF (last accessed Sept. 18, 2020), https://www.nsf.
gov/funding/pgm_summ.jsp?pims_id=505686.
44 See Chapter 8 of this report and its associated Blueprint for Action for its recommendation for
Congress to authorize NIST to sponsor a University Affiliated Research Center (UARC), Federally
Funded Research & Development Center (FFRDC), and/or a lab to provide independent, third-party
testing.
45 The AI4EU project was founded by the European Commission under the H2020 program to
establish the first European AI On-Demand Platform and Ecosystem. The Platform is designed to
support the AI ecosystem and provide a forum to share AI resources from European projects. See
About the Project, AI4EU (last accessed Sept. 18, 2020), https://www.ai4eu.eu/about-project.
46 AI for Humanity, Mila (last accessed Sept, 18, 2020), https://mila.quebec/ia-dans-la-societe/.
47 Artificial Intelligence For Good and For All, Amii (last accessed Sept. 18, 2020), https://www.amii.
ca/.
48 See Pan-Canadian AI Strategy, CIFAR (last accessed Feb. 4, 2021), https://www.cifar.ca/ai; About
Us, Vector Institute for Artificial Intelligence (last accessed Sept. 18, 2020), https://vectorinstitute.ai/
about/.
p
576
BLUEPRINT FOR ACTION: CHAPTER 15
Chapter 15: A Favorable International Technology Order
Annex: Emerging Technology Coalition - Endnotes
49 See Stanford Institute for Human-Centered Artificial Intelligence (last accessed Sept. 18, 2020),
https://hai.stanford.edu/welcome.
50 The U.S. has strong research ties to Israel through the Binational Science Foundation (BSF) and
the Binational Industrial Research & Development Foundation (BIRD). See About the BSF, U.S.-Israel
Binational Science Foundation (last accessed Feb. 3, 2021), https://www.bsf.org.il/about/; What is
BIRD? U.S.-Israel Binational Industrial Research and Development (last accessed Feb. 3, 2021),
https://www.birdf.com/what-is-bird/. The Indo-U.S. Science and Technology Forum (IUSSTF) oversees
the United States-India Science & Technology Endowment Fund (USISTEF), which supports and
fosters joint applied R&D. About the Fund, Indo-U.S. Science and Technology Forum (last accessed
Feb. 3, 2021), https://www.iusstf.org/usistef/us-india-science-technology.
51 See Joint Statement on Artificial Intelligence and Human Rights, Freedom Online Coalition (last
accessed Jan. 5, 2021), https://freedomonlinecoalition.com/wp-content/uploads/2020/11/FOC-Joint-
Statement-on-ArtificiaI-Intelligence-and-Human-Rights.pdf.
52 The FY 2021 NDAA created the Open Technology Fund as Section 309A of the U.S. International
Broadcasting Act of 1994. Pub. L. 116-283, sec. 1299P, William M. (Mac) Thornberry National Defense
Authorization Act for Fiscal Year 2021, 134 Stat. 3388 (2021). Congress has since appropriated
$20 million to the fund. See Joint Explanatory Statement, Division K- Department of State, Foreign
Operations, and Related Programs Appropriations Act, 2021 at 23 (2021), https://www.appropriations.
senate.gov/imo/media/doc/Division%20K%20-%20SFOPS%20Statement%20FY21.pdf (enacted in
Pub. L. 116-260, the Consolidated Appropriations Act, 2021). The Department of State’s Internet
Freedom and Business & Human Rights Section (IFBHR), within DRL, leads U.S. Government policy
and engagement to protect human rights online. See Internet Freedom: Fact Sheet, U.S. Department
of State (Nov. 17, 2017), https://2017-2021.state.gov/internet-freedom/index.html. IFBHR works
across the U.S. Government, with democratic nations, with civil society, and with the Freedom
Online Coalition. IFBHR’s program includes funding development of censorship-defeating peer-to-
peer communications technologies. See Internet Freedom: Advancing and Promoting Peer-to-Peer
internet-freedom-advancing-and-promoting-peer-to-peer-communications-technologies/index.html.
53 Isaac Ben-Israel, et al., Towards Regulation of AI Systems: Global Perspectives on the Development
of a Legal Framework on Artificial Intelligence (AI) Systems Based on the Council of Europe’s
rm.coe.int/prems-107320-gbr-2018-compli-cahai-couv-texte-a4-bat-web/1680a0c17a.
54 See the Chapter 14 Blueprint for Action. See also Guidance on Implementing the UN Guiding
Principles for Transactions Linked to Foreign Government End-Users for Products or Services with
Surveillance Capabilities, U.S. Department of State (last accessed Jan. 4, 2021), https://www.state.
gov/wp-content/uploads/2020/10/DRL-Industry-Guidance-Project-FINAL-1-pager-508-1.pdf.
55 See the Chapter 1 Blueprint for Action on malign information for further details on this proposal.
56 About, Global Internet Forum to Combat Terrorism (last accessed Jan. 5, 2021), https://www.gifct.
org/about/.
57 Civil society in Taiwan has responded to the threat from disinformation in a number of ways,
including demonstrating outside compromised media firms, educating senior citizens on the ways
they may be exposed to disinformation, and the establishment of robust fact-checking groups such
as the Taiwan Fact Check Center, MyGoPen, Cofacts, and Rum Toast. These groups have worked with
both government and social media platforms to not only identify and remove disinformation, but also
to forensically trace disinformation back to sources in China. See Audrey Tang on Taiwan’s Digital
www.stimson.org/2020/interview-with-taiwan-digital-minister-audrey-tang/.
58 For more on democratic use of surveillance technologies, see Chapter 8 of this report, Upholding
Democratic Values: Privacy, Civil Liberties, and Civil Rights in Uses of AI for National Security.
p
577
A FAVORABLE INTERNATIONAL TECHNOLOGY ORDER
59 Kara Frederick, The Razor’s Edge: Liberalizing the Digital Surveillance Ecosystem, CNAS (Sept.
surveillance-ecosystem.
60 See OpenMined’s Efforts for the Coronavirus Pandemic: COVID Alert App, Private Set Interaction, A
Differential Privacy Wrapper and Private Identity, OpenMined (April 1, 2020), https://blog.openmined.
org/openmineds-efforts-for-the-coronavirus-pandemic/.
61 The U.S. and Europe should agree on a common definition for anonymized data to include a clearer
understanding of what constitutes “consent” and “matters of public interest.”
62 Remarks by Angel Gurria, OECD Secretary General, delivered at the 2019 G20 Leaders’ Summit-
Digital (AI, data governance, digital trade, taxation) (June 28, 2019), https://www.oecd.org/g20/
summits/osaka/2019-g20-leaders-summit-digital-osaka-june-2019.htm.
eu/info/sites/info/files/joint-communication-eu-us-agenda_en.pdf.
64 The Commission’s Key Considerations and existing international principles could be leveraged,
such as the OECD Principles on AI, the OECD Guidelines on the Protection of Privacy and
Transborder Flows of Personal Data, the OECD Recommendation on Digital Security of Critical
Activities, the forthcoming OECD Principles on Trusted Government Access to Data, and others.
65 See Chapter 14 of this report and its associated Blueprint for Action for additional details on the
Commission’s recommendations regarding aligning allied export control and investment screening
regimes. Within the U.S. government, the Departments of State and Commerce, on export controls,
and the Departments of State and the Treasury, on investment screening, have already begun such
work.
66 See the Chapter 15 Blueprint for Action.
previous-reports/. In particular, the United States, the Netherlands, and Japan should coordinate
export controls on extreme ultraviolet and ArF immersion lithography tools, as doing so would limit the
ability of China and other competitors to develop the high-end microelectronics that are increasingly
essential for AI. For additional details on the Commission’s recommendations regarding export
controls on SME, see Chapter 14 of this report.
68 See Chapter 13 of this report for additional details and recommendations on the microelectronics
supply chain.
69 See Chapters 3 and 14 of this report.
reports; Andrew Imbrie et al., Agile Alliances: How the United States and Its Allies Can Deliver a
cset.georgetown.edu/research/agile-alliances/; Andrew Imbrie, et al., The Question of Comparative
Advantage in Artificial Intelligence: Enduring Strengths and Emerging Challenges for the United
wp-content/uploads/CSET-The-Question-of-Comparative-Advantage-in-Artificial-Intelligence-1.pdf.
reports/.
72 See Chapters 13 and 16 of this report.
73 “IP5” is the name of the forum of the world’s five largest IP offices that was set up to improve
the efficiency of the examination process for patents worldwide. See About IP5 Co-Operation,
fiveIPoffices (last accessed Jan. 4, 2021), https://www.fiveipoffices.org/about.
74 See Chapter 12 of this report and its associated Blueprint for Action.
p
578
BLUEPRINT FOR ACTION: CHAPTER 15
Chapter 15: A Favorable International Technology Order
Annex: Emerging Technology Coalition - Endnotes
75 See Chapter 12 of this report and its associated Blueprint for Action.
76 See Chapter 10 of this report.
77 This approach has been recommended by Melissa Flagg and Zachary Arnold. See Melissa Flagg
& Zachary Arnold, A New Institutional Approach to Research Security in the United States Defending
georgetown.edu/research/a-new-institutional-approach-to-research-security-in-the-united-states/.
78 See Chapter 1 and its associated Blueprint for Action on preparing for AI-enabled cyber conflict.
p
579
BLUEPRINT FOR ACTION: CHAPTER 16
Chapter 16:
Associated Technologies
Blueprint for Action
Recognizing that leadership in artificial intelligence (AI) relies on leadership across a suite
of emerging technologies, the United States must prioritize the research and development
(R&D), application, and adoption of not just AI, but the technologies that enable it and are
enabled by it. This process should be based on a careful analysis of the national security
threats and opportunities at the intersection of AI and its associated technologies. If the
U.S. government fails to adopt a more strategic approach to protecting and promoting U.S.
advantages in these areas, it risks jeopardizing the country’s technological leadership,
economic prosperity, and national security.
In accordance with its mandate to consider both AI and “associated technologies,”
the Commission identifies and proposes steps to maintain U.S. leadership across the
spectrum of technologies it believes are most critical to U.S. national competitiveness. The
Commission then offers specific recommendations on how the United States can proactively
address the novel national security threats and opportunities posed by three technologies
in particular: biotechnology, quantum computing, and 5G telecommunications.1 Finally,
the Commission expands its analysis to include recommendations on a broader set of
emerging technologies critical to U.S. national competitiveness.
Technologies Critical to U.S. National Competitiveness
The Commission has identified eight technologies and related platforms that are key to
U.S. leadership. Two of these technologies—AI and microelectronics—are addressed
elsewhere in this report. The remaining six—biotechnology, quantum computing, 5G and
advanced networking, autonomy and robotics, advanced and additive manufacturing, and
energy systems—are covered below. These recommendations build on the Commission’s
previous work by providing actions the U.S. government could take to promote overall
U.S. leadership and long-term competitiveness across the constellation of emerging
technologies.
Recommendation: Identify and Prioritize Technologies Central to National Competitiveness
Recommendation
To date, there is no whole-of-government consensus for which emerging technologies
are most critical to long-term strategic competitiveness and whose development must
p
581
ASSOCIATED TECHNOLOGIES
be prioritized. Several government agencies have made independent attempts to define
such a list: the 2018 National Defense Strategy,2 the list of “critical emerging technologies”
produced by the Department of Defense in response to Section 1793 of the FY 2019
National Defense Authorization Act (NDAA),3 the Department of Commerce’s 2018 Advance
Notice of Proposed Rulemaking (ANPRM) of controls on certain emerging technologies,4
the report by the President’s Council of Advisors on Science and Technology titled
Recommendations for Strengthening American Leadership in Industries of the Future from
2020,5 and the bill introduced by Senator Chuck Schumer in the 116th Congress with seven
bipartisan co-sponsors titled the “Endless Frontier Act.”6 Additionally, the White House
published the National Strategy for Critical and Emerging Technologies in October 2020,
which included a list of critical and emerging technologies.7 However, this document does
not explain why each of these technologies is essential to U.S. national competitiveness,
nor does it include specific implementation plans for promoting their development and
protecting U.S. advantages in each.
These lists have substantial overlap, but no two lists are the same and no single list is
authoritative. Consequently, there is no whole-of-government consensus, and certainly no
national consensus, of which technologies are critical to U.S. national competitiveness,
making it more difficult for the U.S. government to marshal private-sector investment, for
legislators to prioritize funding, and for U.S. government agencies to coordinate technology
protection and promotion. There is also no list around which the White House can organize
a national technology strategy and no coordinated mechanism within the U.S. government
to support financing of these priorities when there are market failures and private-sector
financing is insufficient.
Actions for the Executive Office of the President:
• Define and prioritize the key emerging technologies in which U.S. leadership is
essential.
o The Executive Office of the President, in consultation with departments and
agencies, should publish a single, authoritative list of technologies and sectors
which are key to overall U.S. competitiveness, along with detailed implementation
plans for each to ensure long-term U.S. leadership.
■ The implementation plans should identify specific subcomponents of each
technology that are most important, key choke points where competitors
could be blocked with minimal impact on U.S. industry, and where additional
resources are needed. These plans should include specific steps to
promote domestic industry, ensure supply chain resiliency, and protect key
technologies from competitors. This list of technologies and the associated
implementation plans will form the core of a National Technology Strategy, as
referenced in Chapter 9 of this report.
o The creation and maintenance of such a list and implementation plans will help
produce a national consensus regarding which industries are most important in the
emerging techno-economic competition. The result will be an important message
to Congress regarding where the country must prioritize and expend resources,
p
582
BLUEPRINT FOR ACTION: CHAPTER 16
as well as a powerful demand signal to industry. The figure below includes eight
technologies that the Commission recommends be considered for the list.
U.S. Government Lists of Critical Technologies
NSCAI-
2018
DoD List
Commerce
PCAST List
S.3832 -
WH Nat
Proposed
National
of Critical
ANPRM on
of Industries
Endless
Strategy
Critical
Defense
Emerging
Emerging
of the Future
Frontier
for
Technology List
Strategy
Technologies
Technologies
Act
C&ET
Artificial
Intelligence
Biotechnology
Quantum
Computing
Semiconductors
and Advanced
Hardware
Autonomy and
Robotics
5G and
Advanced
Networking
Advanced
Manufacturing
Energy
Systems
• Expand the loan authority of the Development Finance Corporation to include
domestic industrial base capabilities supporting key emerging technologies.
o The President should issue an Executive Order that expands the loan authority of
the U.S. International Development Finance Corporation (DFC) to include domestic
industrial base capabilities related to any of the aforementioned technologies that are
identified by the Executive Office of the President as key to overall U.S. competitiveness.
■ Specifically, the Executive Order should delegate authority under Title III of
the Defense Production Act to the DFC to issue loans that “create, maintain,
protect, expand, or restore domestic industrial base capabilities” supporting the
aforementioned list of technologies, or “the resiliency of relevant domestic supply
chains.” This new authority should be of indefinite duration.
■ This action would build off of Executive Order 13922, which expanded similar
domestic loan authorities to DFC related to industries supporting “the national
response and recovery to the COVID-19 outbreak” until 2022.8
p
583
ASSOCIATED TECHNOLOGIES
o Expanding the domestic authorities of the DFC as it relates to critical technologies
will help the government support key platforms and projects which are critical
to future U.S. national security and economic competitiveness but lack sufficient
private-sector capital.
■ The DFC should coordinate with the Technology Competitiveness Council
recommended in Chapter 9 of this report to identify specific platforms that are
most in need of such financing.
Ensuring U.S. Leadership in Biotechnology
The combination of advances in AI and biology have the potential to reshape the global
economy for the next century. Progress in genetic sequencing has given researchers the
ability to read the “code of life.” Given the significant quantity of data involved, AI will
be essential to fully understanding how genetic code interacts with biological processes.
Finally, advances in synthetic biology and genetic editing will give researchers the ability
to manipulate this code to perform specific functions. Together, these techniques will
enable transformational breakthroughs in biology and underpin most future scientific
breakthroughs related to human health, agriculture, and climate science. The nation
which is best able to simultaneously leverage both technologies will have substantial
strategic advantages for the foreseeable future, potentially becoming a global leader
in pharmaceuticals, reducing its reliance on foreign supply chains, and even ensuring
it has a healthier and more capable population. These technological breakthroughs will
also cause the biotechnology sector to become a major driver of overall U.S. economic
competitiveness.
Recommendation: Prioritize the Development of an Advanced Biotechnology R&D
Recommendation
Ecosystem
The United States must invest in key platforms that better position the U.S. academic and
commercial biotech industry to benefit from AI-enabled advancements in biology. It should
specifically look to support platforms that aggregate biodata, and specifically genetic data,
in a secure manner in order to enhance the ability of U.S. researchers to utilize AI to facilitate
breakthrough biotechnology research and innovation. Additionally, the United States should
support efforts to expand the scope and sophistication of U.S. biofabrication capabilities
to ensure it can keep pace with forthcoming research advancements. It should specifically
support efforts to transform the biotechnology industry away from its current, vertically
integrated models and encourage the development of multiple standardized, merchant
biofabrication facilities. Doing so would expand access to advanced biofabrication tools
among startups and laboratories by allowing firms to rapidly design new molecules and
materials via the cloud and place immediate orders for fabrication.
p
584
BLUEPRINT FOR ACTION: CHAPTER 16
Actions for Congress and the Department of Health and Human Services:
• Fund and establish a world-class biobank for genetic data.
o Congress should fund efforts to build a world-class biobank within the National
Institutes of Health (NIH). The current leading U.S. genetic database, GenBank, is
underfunded, difficult to access, and poorly curated, particularly in comparison to
other leading genetic databases such as the U.K. BioBank or the China National
GeneBank. The entity should be securely and easily accessible by legitimate
researchers; contain a wide variety of whole human, animal, and plant genomes,
including de-identified metadata about phenotypes; and aggregate other open and
potentially even proprietary datasets for specialized uses. It must also include strong
privacy protections for human genetic data. Creating and staffing such an entity
would likely require a budget of approximately $100 million per year, on top of up-
front construction costs.9
• Direct funding to support advanced biotech manufacturing initiatives through
entities such as BARDA.
o The Department of Health and Human Services should direct funds to support
advanced biotech manufacturing initiatives through entities such as the Biomedical
Advanced Research and Development Authority (BARDA), and Congress should
prioritize such initiatives in future health-related spending bills. This could take
the form of financial incentives for advanced biotech manufacturing firms focused
on sophisticated, flexible, cloud-based fabrication, or R&D funding to support
advanced manufacturing techniques.
Recommendation
Recommendation: Prioritize Advanced Biotechnology Capabilities as Imperative for
National Security and Economic Competitiveness
The growing importance of biotechnology leadership to health, food, production, and
science also makes it a national security imperative that the United States take proactive
steps to facilitate long-term U.S. leadership in the field. Advancements in biotechnology
will also create novel national security challenges, ranging from engineered pathogens
to augmented competitor human physiological or mental capabilities. The United States
currently is not postured to address such challenges, and biological threats have rarely
been a priority issue for the U.S. national security community. The COVID-19 pandemic
clearly illustrates that the United States must think more broadly about national security
threats than it has in the past, and that biological threats in particular have the potential to
impose significant costs on U.S. society and security.
U.S. competitors see the potential for AI to spur new, transformational advances in
biotechnology. China in particular is actively seeking global leadership in both fields, sees
its AI and biotechnology strategies as mutually reinforcing, and believes the synergies
between the two will translate into military advantage.10 China also faces fewer barriers to
collecting, using, and combining human biological data given its disregard for individual
privacy and bioethical principles. The global reach of China’s genomics giant, BGI, poses
similar threats in the biotechnology sector as Huawei does in the communications sector.
p
585
ASSOCIATED TECHNOLOGIES
Actions for the Executive Office of the President:
•
Update the U.S. National Biodefense Strategy to include additional AI-enabled
biological threats.11
o The National Security Council should update its National Biodefense Strategy,
which currently only focuses on natural or engineered pathogens, to include a
wider vision of biological threats.12 The strategy should specifically examine how AI
could enable new biological advances which pose unique national security threats,
such as human enhancement, and how U.S. competitors could utilize advantages
in biotechnology or biodata as an instrument of national power. It should also
specifically consider how AI could identify and counter the creation of advanced,
engineered pathogens which target certain elements of the U.S. population or food
supply. AI is facilitating a rapid evolution of the biotechnology field, and the U.S.
biodefense strategy must evolve with it.
•
Direct departments and agencies to prioritize initiatives that promote U.S.
biotechnology leadership.
o Directing departments and agencies to prioritize initiatives promoting U.S.
biotechnology leadership would include aggressively promoting funding for basic
research in biology, particularly applications of biology that utilize AI; focusing
resources on forecasting how AI will enable future biotechnology breakthroughs;
and continuing to cultivate talent both inside and outside the government, as well
as commercial activity at the nexus of AI and biology. This will require an entity
which is empowered to coordinate across the economic, technological, and security
spheres, such as the Commission’s recommended Technology Competitiveness
Council.13
Recommendation: Publicly Highlight BGI’s Links to the Chinese Government
Recommendation
BGI is China’s de facto national champion in genetic sequencing and research and is
among the world leaders in DNA sequencing. It has research affiliations with multiple U.S.
universities, including the University of Washington and Washington State University.14 BGI
has also benefited from substantial support from the Chinese government, as well as its
2013 acquisition of a competing U.S. firm, Complete Genomics.15 There are indications
that BGI’s links with the Chinese government may run deeper than it publicly claims, as it
built and operates China National GeneBank, the Chinese government’s national genetic
database, and has used PLA-owned supercomputers to process genetic information.16
Chinese diplomats have pushed BGI-built COVID-19 testing kits, including in the United
States, and by August 2020 BGI had “sold 35 million rapid COVID-19 testing kits to 180
countries, and built 58 labs in 18 countries.”17
BGI may be serving, wittingly or unwittingly, as a global collection mechanism for Chinese
government genetic databases, providing China with greater raw numbers and diversity
of human genome samples as well as access to sensitive personal information about key
individuals around the world. The highest levels of the United States government should
publicly state these concerns so as to raise awareness among the U.S. commercial and
p
586
BLUEPRINT FOR ACTION: CHAPTER 16
academic biotechnology communities, as well as U.S. allies, many of which currently have
partnerships or business dealings with BGI.
Action for the Department of State:
• Launch a strategic communications campaign to publicly highlight the links
between the Chinese government and BGI.
o The Secretary of State should personally voice concern about BGI’s ties to
the Chinese government and instruct the Department to conduct a strategic
communications campaign to highlight those links and warn of the dangers of
the Chinese government obtaining personal genetic information via BGI. The
Department should also warn BGI and the Chinese government that it will closely
monitor BGI’s activities, and that should BGI be utilized as a mass DNA-collection
apparatus for the Chinese government it could face additional U.S. regulatory
action.
Recommendation: Pursue Global Cooperation on Smart Disease Monitoring
Recommendation
While pivoting to a more competitive national approach toward biotechnology policy,
the United States should also pursue efforts to enhance global cooperation on disease
monitoring. By pooling existing open-source health-related data with improved early warning
signals and data on zoonotic spillovers and transmission of novel viruses, governments will
be better postured to use AI to predict and contain future pandemics. Combining increased
transparency and data sharing on disease outbreaks with AI tools—which can enhance
early outbreak detection and contribute to real-time disease monitoring—could provide
substantial benefit for global public health if all countries, including China, participated in
good faith.18
Action for the Departments of State and Health and Human Services:
• Support multilateral efforts to promote smart disease monitoring.
o The Departments of State and Health and Human Services should lead and support
multilateral efforts to promote smart disease monitoring. In particular, the United
States should pursue efforts to integrate and standardize international health-
related data sets and combine them with global data about zoonotic spillovers
to allow for the utilization of AI technologies to create shared, predictive, global
disease-monitoring tools.
Ensuring U.S. Leadership in Quantum Computing
Quantum computing has the potential to create new national security threats and
opportunities by enhancing the speed and precision of existing AI systems and creating
new capabilities that could fundamentally alter the strategic environment. For example,
quantum computers may be able to more efficiently optimize logistics for the military or
p
587
ASSOCIATED TECHNOLOGIES
discover new materials for weapon systems.19 Quantum sensors and communications
are also poised to revolutionize the collection and transfer of sensitive information, which
directly affects how AI is trained and deployed in national security use cases.20 Failure
to step up investment in the R&D of materials and components for quantum computers,
open-source software tools, and hybrid quantum-classical algorithms that leverage noisy
intermediate-scale quantum computers may leave the United States vulnerable to strategic
surprise on behalf of competitors.21
Recommendation: Transition from Basic Research to National Security Applications of
Recommendation
Quantum Computing
Although the United States is well-positioned to take advantage of its early success in
the basic science of quantum computing, the U.S. Government must increase its focus
on fielding national security applications or risk falling behind strategic competitors.
Most notably, China has made significant investments in military applications of quantum
computing in an attempt to offset U.S. strengths.22 The Department of Defense (DoD) is still
refining its approach to rapidly transition commercial technologies from research to fielding
in high-cost, hardware-intensive sectors such as quantum computing. In the long term,
DoD should prioritize efforts to rapidly procure technology across its innovation offices, but
this process could take several years of dedicated effort. In the interim, announcements
of priority applications will help spur private-sector investment and innovation in quantum
computing despite the absence of an integrated technology-procurement apparatus.23
Action for the President:
• The President should direct departments and agencies to announce priority use
cases of quantum computers.
o The National Quantum Coordination Office (NQCO) should coordinate an effort by
departments and agencies represented on the National Science and Technology
Council (NSTC) Subcommittee on Quantum Information Science (QIS) to announce
their priority use cases of quantum computers. By reflecting the combined views
of federal entities engaging with the private sector, this would signal that a market
for practical applications of quantum computing exists, set clear and specific goals
for the private sector to pursue, and incentivize additional private investment.
Some applications of quantum computers may be too sensitive to reveal publicly,
but those that can be announced will provide direction to the private sector and
facilitate the commercialization of quantum computers, which can then be applied
to national security use cases.
Recommendation: Foster a Vibrant Domestic Quantum Fabrication Ecosystem
Recommendation
Due to the strategic implications of quantum computing and its application to AI, the
United States must take steps now to cement its long-term status as the global leader in
the design and manufacturing of quantum processing units (QPUs). To avoid the situation
in which the U.S. semiconductor industry currently finds itself, the United States must
p
588
BLUEPRINT FOR ACTION: CHAPTER 16
establish trusted and assured sources for critical materials and components of QPUs,
ranging from manufacturing equipment to superconductors and dilution refrigerators.24
Although these materials and components may not yet represent choke points, they will
inevitably become more specialized as the manufacturing processes required to design
and produce QPUs continue to advance. Rather than reshoring the entire supply chain
for QPUs, the United States should work with its allies to develop a resilient network of
suppliers for critical components that directly impact U.S. national security.
However, a secure supply chain is not sufficient to ensure U.S. leadership in quantum
computing. To benefit from future breakthroughs in the field, the United States must create
a robust domestic ecosystem for the research, development, and application of quantum
computers that attracts top-tier talent from around the world.25 The U.S. Government
should offer incentives for the R&D of quantum computers and their components
while simultaneously creating demand for national security applications of quantum
technologies. The Quantum Economic Development Consortium (QED-C), proposed in
the National Quantum Initiative (NQI) Act of 2018, is an important step toward extending
U.S. leadership in next-generation computer hardware for years to come.26
Action for Congress:
• Enact a package of provisions that incentivizes the domestic design and
manufacturing of quantum computers and their constituent materials.
o A tax credit for expenditures made in the United States on research and
development, manufacturing equipment, and workforce training related to the
development of quantum computers is a necessary, albeit not sufficient, step to
maintain U.S. competitiveness in this area. This provision could be modeled on the
Alternative Simplified Credit (ASC), which provides a credit of 14% of expenditures
on R&D in excess of 50% of base period expenditures. To help startups on the
cutting edge of research and development access funding that allows them
to scale, the U.S. Government should also provide loan guarantees and equity
financing.
Recommendation
Recommendation: Make Quantum Computing Accessible to Researchers via the National
AI Research Resource (NAIRR)
Despite recent advances in the fields of quantum hardware and software, fault-tolerant
quantum computers (FTQCs) capable of performing general-purpose tasks are unlikely
to replace classical computers anytime soon. In the near term, the United States should
invest in noisy intermediate-scale quantum (NISQ) computers that are capable of deriving
probabilistic solutions from imperfect qubits.27 Hybrid quantum-classical techniques have
also shown promise, whereby classical computers delegate certain tasks to purpose-built
quantum devices within the same workflow. However, resources suitable for developing this
type of software are not readily accessible.28 By making classical and quantum computers
available in the same workflow, the U.S. Government would lower barriers to innovation for
p
589
ASSOCIATED TECHNOLOGIES
startups in the quantum computing space and attract top-tier talent from around the world.
The resulting public-private partnerships would also encourage the commercialization
of quantum computers and help the U.S. Government adopt those products for national
security use cases.
Action for the Executive Branch:
• Make classical and quantum computers available in the same workflow via the
National AI Research Resource.
o By providing access to both classical and quantum computers via the National AI
Research Resource (NAIRR), which the Commission recommended establishing in
its First Quarter Recommendations and describes in greater detail in Chapter 11 of
this report,29 the U.S. Government would help researchers from industry, academia,
and government build and test sofware tools and algorithms that leverage both
classical and quantum computers in a hybrid fashion. These types of applications
are likely to be the nearest-term use case of quantum computers.
Ensuring U.S. Leadership in 5G Telecommunications
AI systems require high-fidelity sensing as well as fast, safe, and secure networks. It is a
national security imperative for the U.S. military and the nation as a whole to have access
to a powerful 5G network to enable future AI capabilities and ensure the network is trusted.
The United States must preserve this access and trust while building out commercial 5G
networks domestically and internationally.
Recommendation: Accelerate U.S. 5G Deployment Through Spectrum Sharing
Recommendation
The slow rollout of 5G networks in the United States compared to China risks undermining
U.S. advances in AI, both in the government and the private sector.30 The sub-6 GHz
spectrum, sometimes referred to as the mid-band or the “goldilocks” band of spectrum,
is the critical portion of the spectrum for both DoD and commercial 5G operations. Sub-
6 GHz spectrum is critical for 5G civilian communications since it combines high data
rates with good range and penetration. Within DoD, it is also already used by many radar
and communication systems because it also combines high discrimination capability
with long-range operations.31 In part due to its importance to military operations, DoD
has retained exclusive access to significant portions of the mid-band spectrum, which
limits commercial uses. Unfortunately, the lack of U.S. mid-band spectrum commercial
availability is substantially slowing the deployment of 5G networks domestically. Given that
sub-6 GHz is important for sensing using radar and civilian communications, spectrum
sharing between DoD and the private sector is the ideal approach to enabling access for
both purposes in a manner that balances national security and economic interests.32
Several U.S. Government agencies are working to address this problem by developing
spectrum-sharing capabilities within the
3- to
6-GHz range. In
2015, the Federal
p
590
BLUEPRINT FOR ACTION: CHAPTER 16
Communications Commission
(FCC) established the Citizens Broadband Radio
Service (CBRS), the first U.S. spectrum sharing model.33 Since that time, the National
Telecommunications and Information Administration
(NTIA) has studied, and has
collaborated with the DoD and FCC on, maximizing spectrum-sharing capabilities.34
The CBRS enables shared federal and non-federal use of the band. This work allows
the U.S. Navy and non-government providers to share the 3550-3700 MHz band across
three dynamically managed tiers: the Navy will maintain first priority access, followed
by companies and organizations that purchase priority-access licenses, and finally
companies and organizations that register at no cost. The FCC held its first auction for
priority-access licenses for this band in July 2020, which raised more than $4.5 billion
through the sale of 20,625 licenses.35 This is a promising but modest start and these
efforts must expand to a larger portion of the mid-band spectrum to be competitive with
China. To achieve spectrum sharing at a competitive level will require technical analysis
and engagement with industry. A comprehensive process will be critical to ensuring that
DoD maintains access to spectrum essential for operational effectiveness while also
broadening commercial access to spectrum for civilian 5G networks.36
Action for the NTIA, FCC, and DoD:
• Expand spectrum-sharing programs led by NTIA, FCC, and DoD, starting with a
one-year 5G spectrum-sharing demonstration program.
o The Commission urges NTIA, the FCC, and DoD to jointly expand spectrum-sharing
programs such as the CBRS and work to license additional sub-6GHz spectrum
to wireless carriers and equipment makers for commercial 5G use. Sharing and
licensing additional mid-band spectrum will ensure unrestricted DoD access in the
event of an emergency while also opening up 5G for commercial use. However,
current spectrum-sharing capabilities must be further analyzed, tested, and
demonstrated before they can be scaled. The Commission supports a one-year
demonstration program that includes NTIA, FCC, DoD, and industry to assess the
network’s capabilities and its capacity to dynamically share spectrum between
government and civilian users. If successful, such a network would be rapidly scaled
with commercially available equipment.
Promote U.S. Leadership in Other Key Emerging Technologies
AI, microelectronics, biotechnology, quantum computing, and 5G telecommunications are
not the only emerging technologies that will underpin U.S. national competitiveness in the
21st century. The Commission assesses that the full spectrum of emerging technologies
key to U.S. technological leadership extends further and includes autonomy and robotics,
advanced manufacturing, and energy systems. The Commission therefore recommends
several actions to ensure U.S. leadership in these additional key emerging technologies.37
p
591
ASSOCIATED TECHNOLOGIES
Autonomy and Robotics
Recommendation: Incentivize the Development of World-Class Software Platforms for
Recommendation
Robotic and Autonomous Systems
Autonomous systems that rely on robotics to execute tasks in the real world are being applied
to everything from advanced manufacturing to warfighting.38 As AI continues to improve
the ability of these systems to match or exceed human capabilities, the United States
must position itself as a leading producer and adopter of robotic hardware and software
for civilian and military use cases. The United States currently lags behind countries such
as Japan and Korea on the manufacturing and installation of industrial robots, and China
has declared robotics as a core industry.39 As the United States reshores certain strategic
supply chains and increases its reliance on autonomous systems, continued access to
cutting-edge robotics will be a national security imperative.
Action for the National Institute of Standards and Technology:
• Incentivize the development of world-class sofware platforms for robotic systems
by U.S. firms.
o By designing the sofware platforms upon which core robotic capabilities are built,
U.S. firms will be well-positioned to shape the next wave of industrialization. The
U.S. government should expand collaboration with industry on basic R&D, set
international standards, and share data pertaining to robotic system development
by expanding upon the work of the Intelligent Systems Division at the National
Institute of Standards and Technology (NIST).40 The U.S. government should
also incentivize the early adoption of robotic systems across the public and private
sectors by creating markets in areas ripe for automation.41 These efforts will yield
valuable data and experience in scaling automation and facilitate the application
of robotics to adjacent sectors. A multipronged approach along these lines will
position U.S. industry to compete more effectively in the market for robotic systems
sofware, a strategically important area that is compatible with existing U.S.
strengths.
Advanced Manufacturing
Recommendation: Accelerate Additive Manufacturing Production of Legacy Parts Across
Recommendation
the Department of Defense
The ability to manufacture high-tech products domestically is critical to a nation’s security
and its economic productivity. The United States must strive to develop manufacturing
capabilities in industries that are essential to crisis response or that would take too long
to bring online in the event of a protracted conflict.42 Innovation also benefits from the
co-location of firms engaged in technological design and those that produce finished
products, which enables rapid feedback and continuous iteration on product design.43
This link is particularly important in the defense sector, where communication between
p
592
BLUEPRINT FOR ACTION: CHAPTER 16
researchers, designers, and manufacturers can help quickly transition a technology from
the lab to the field. However, the United States has relinquished manufacturing leadership
in high-tech industries that employ highly skilled workers to high-wage nations like Germany
and Japan.44 Meanwhile, China and other lower-wage nations are moving up the value
chain from low-value manufacturing processes, such as assembly, to more sophisticated
techniques.45 Although the supply chain disruptions resulting from the COVID-19 pandemic
may prompt the return of some manufacturing to the United States, the broader trend of
offshoring the manufacturing of next-generation technologies appears likely to continue
unless the U.S. government takes appropriate action.46
Action for the Department of Defense:
• Accelerate additive manufacturing of legacy parts across the Department of
Defense.
o Additive manufacturing and 3D printing have the potential to transform the
manufacturing industry by enabling the rapid production of complex objects on
demand and at the point of need.47 Although existing 3D printers cannot match
the quality of advanced traditional techniques, AI has shown the potential to
significantly improve the accuracy of 3D printing.48 The DoD should proactively
support the improvement of 3D printing by identifying all legacy parts in active
weapon systems suited to production by additive manufacturing and 3D printers
and commit to doing so by 2025.49
Energy Systems
Recommendation
Recommendation: Develop and Domestically Manufacture Energy Storage Technologies
to Meet U.S. Market Demand by 2030
Cheap and reliable access to energy is critical to U.S. national security. Although the United
States is at the forefront of the exploration, extraction, and processing of oil and gas and
possesses significant domestic reserves, China is by far and away the leading producer
of renewable energy and is investing heavily in advanced energy storage technologies,
such as batteries and their constituent materials.50 As the cost of intermittent renewable
sources continues to fall, the United States must commit to developing and deploying the
next generation of energy storage devices, from long-duration stationary applications to
battery packs for electric vehicles.
Action for Congress:
• Fund the Department of Energy’s initiative to develop and domestically
manufacture energy storage technologies to meet U.S. market demand by 2030.
o Improving the cost and energy density of storage technologies will drive progress
in sectors ranging from electric vehicles to distributed energy generation. The
Department of Energy (DoE) has set the ambitious goal of developing and
domestically manufacturing storage technologies capable of meeting the entirety
p
593
ASSOCIATED TECHNOLOGIES
of U.S. market demand by 2030.51 Congress should fully fund the federal R&D
needed to achieve the DoE’s Energy Storage Grand Challenge roadmap by 2030
and establish appropriate incentives for the commercialization of the resulting
technologies.52
Further Consideration of Additional Technologies and Conclusion
While the Commission believes the eight emerging technologies discussed above and
elsewhere in this report—AI, microelectronics, biotechnology, quantum computing, 5G
telecommunications, autonomy and robotics, advanced manufacturing, and energy
systems—will be crucial to future national competitiveness, this list is by no means
exhaustive. Other emerging technologies and platforms—everything from digital currencies
and other types of financial technology to space systems—will likely also play a major role
in the U.S. economy and its national security moving forward. And there are undoubtedly
technologies that have yet to be created which, in the near future, will have transformative
effects on the lives and security of American citizens.
We are at the beginning of a new era, in which technologies not only are the principal
driver of global markets and geopolitics, but they also advance and emerge faster than
ever before. As the speed of technological development accelerates and an increasing
number of technologies have dual-use applications, techno-national security threats will
continue to multiply. To meet this challenge, the U.S. government must continually assess
new technological advancements to determine their potential to disrupt industries, change
economies, and transform national security.
The process of technology horizon-scanning, forecasting, and proactively crafting
policies to address upcoming national security threats related to emerging technologies
must become an ingrained component of the U.S. national security process. Doing so
is not only essential, but also urgent. If the U.S. government waits to adapt to this new
reality until a subsequent commission makes a similar recommendation, it will likely be
playing technological catch-up from a position of national security weakness. As existing
technologies evolve and new ones emerge, the relationship between technology and
national security will only grow stronger, and the need for the United States to maintain
overall technical leadership will only increase.
p
594
BLUEPRINT FOR ACTION: CHAPTER 16
Blueprint for Action: Chapter 16 - Endnotes
1 The Commission identified these as essential to overall U.S. technological leadership in its 2019
Interim Report. See Interim Report, NSCAI at 31 (Nov. 2019), https://www.nscai.gov/previous-reports/.
2 Summary of the 2018 National Defense Strategy of the United States of America, U.S. Department
of Defense at 3 (2018), https://dod.defense.gov/Portals/1/Documents/pubs/2018-National-Defense-
Strategy-Summary.pdf.
3 Fiscal Year 2019 Industrial Capabilities: Report to Congress, U.S. Department of Defense at 132
20%20RPT%20Subj%20FY19%20ICR%2007092020.pdf?ver=2020-07-10-124452-180.
4 83 Fed. Reg. 58201, Review of Controls for Certain Technologies, U.S. Department of
Commerce: Bureau of Industry and Security (Nov. 19, 2018), https://www.federalregister.gov/
documents/2018/11/19/2018-25221/review-of-controls-for-certain-emerging-technologies.
5 Recommendations for Strengthening American Leadership in Industries of the Future, President’s
Council of Advisors on Science and Technology (June 2020), https://science.osti.gov/-/media/_/pdf/
about/pcast/202006/PCAST_June_2020_Report.pdf.
6 S. 3832, 116th Cong. (2020), https://www.congress.gov/bill/116th-congress/senate-bill/3832.
7 National Strategy for Critical and Emerging Technologies, The White House at A-1 (Oct. 2020),
https://trumpwhitehouse.archives.gov/wp-content/uploads/2020/10/National-Strategy-for-CET.pdf.
8 Donald J. Trump, Executive Order 13922: Delegating Authority Under the Defense Production Act
to the Chief Executive Officer of the United States International Development Finance Corporation To
gov/documents/2020/05/19/2020-10953/delegating-authority-under-the-defense-production-act-to-
the-chief-executive-officer-of-the-united.
9 For comparison, the Chinese government provided approximately $117 million in initial funding to the
China National GeneBank for its construction and creation. See Zhuang Pinghui, China Opens First
National Gene Bank, Aiming to House Hundreds of Millions of Samples, South China Morning Post
gene-bank-opens-shenzhen.
10 Elsa Kania, Minds at War: China’s Pursuit of Military Advantage Through Cognitive Science and
prism_8-3_Kania_82-101.pdf.
11 This recommendation is included in Chapter 1 of this report.
content/uploads/2018/09/National-Biodefense-Strategy.pdf.
13 See Chapter 9 of this report for additional details on the proposed Technology Competitiveness
Council.
14 See, e.g., BGI & US Collaborate on Precision Medicine Development, UW Medicine (May 10, 2016),
https://newsroom.uw.edu/story/bgi-uw-collaborate-precision-medicine-development.
15 In 2010, BGI received a $1.5 billion loan from the state-run China Development Bank. The precise
extent of government subsidies to BGI are unknown, but likely substantial. See Kirsty Needham,
new-doors-for-chinas-gene-giant-idUSKCN2511CE; see also Antonio Regaldo, China’s BGI Says
technologyreview.com/2020/02/26/905658/china-bgi-100-dollar-genome/.
careers/china-national-genebank-officially-opens/.
p
595
ASSOCIATED TECHNOLOGIES
17 See Kirsty Needham, Special Report: COVID Opens New Doors for China’s Gene Giant, Reuters
report-covid-opens-new-doors-for-chinas-gene-giant-idUSKCN2511CE; see also Jeanne Whalen &
Elizabeth Dwoskin, California Rejected Chinese Company’s Push to Help with Coronavirus Testing.
business/2020/07/02/china-bgi-california-testing/.
18 See Chapters 9 and 15 of this report for additional information on cooperation on issues at the
intersection of AI and global health.
19 Pontus Vikstål, et al., Applying the Quantum Approximate Optimization Algorithm to the Tail-
Assignment Problem, Physical Review Applied Vol. 14, Iss. 3 (Sept. 3, 2020), https://doi.org/10.1103/
PhysRevApplied.14.034009; He Ma, et al., Quantum Simulations of Materials on Near-Term Quantum
Computers, npj Computational Materials (July 2, 2020), https://doi.org/10.1038/s41524-020-00353-z.
20 C.L. Degen, et al., Quantum Sensing, arXiv (June 7, 2017), https://arxiv.org/pdf/1611.02427.pdf;
Juan Yin, et al., Entanglement-Based Secure Quantum Cryptography over 1,120 Kilometres, Nature
582, 501-505 (June 15, 2020), https://doi.org/10.1038/s41586-020-2401-y.
21 In December 2020, a team of researchers in China demonstrated quantum advantage on a photonic
quantum computer. See Han-Sen Zhong, et al., Quantum Computational Advantage Using Photons,
Science (Dec. 18, 2020), https://science.sciencemag.org/content/370/6523/1460.
22 Elsa B. Kania & John Costello, Quantum Hegemony? China’s Ambitions and the Challenge to U.S.
hegemony.
23 For additional information and NSCAI views on quantum computing, AI, and national security, see
nscai.gov/previous-reports/.
24 Applications of Quantum Technologies: Executive Summary, Defense Science Board at C-1
Summary_10.23.2019_SR.pdf.
25 Dario Gil, How to Ensure the U.S.’s Quantum Future, Scientific American (Aug. 20, 2020), https://
www.scientificamerican.com/article/how-to-ensure-the-uss-quantum-future/.
26 The bipartisan National Quantum Initiative Act of 2018 prompted a series of steps that establish
quantum computing as a strategic priority for the United States. The Department of Energy (DoE)
announced $625 million to establish five Quantum Information Science research centers over
five years led by the national laboratories. The National Science Foundation (NSF) announced
$75 million to create three Quantum Leap Challenge Institutes over the same period. Lastly,
the President’s FY 2021 Budget recommended doubling federal investment levels in quantum
technologies by 2022. This continuing investment is necessary to determine the full potential of
quantum computing and maintain the United States’ position of leadership in next-generation
computer hardware. For more details, see Pub. L. 115-368, National Quantum Initiative Act, 115th
Cong. (2018), https://www.congress.gov/bill/115th-congress/house-bill/6227; Department of
Energy Announces $625 Million for New Quantum Centers, U.S. Department of Energy (Jan. 10,
centers; NSF Establishes 3 New Institutes to Address Critical Challenges in Quantum Information
announcements/072120.jsp; Recommendations for Strengthening American Leadership in
Industries of the Future, The President’s Council of Advisors on Science and Technology at 13 (June
pdf?la=en&hash=019A4F17C79FDEE5005C51D3D6CAC81FB31E3ABC.
27 John Preskill, Quantum Computing in the NISQ Era and Beyond arXiv at 4, 14 (July 30, 2018),
https://arxiv.org/pdf/1801.00862.pdf.
p
596
BLUEPRINT FOR ACTION: CHAPTER 16
Blueprint for Action: Chapter 16 - Endnotes
28 The Department of Energy and the Air Force offer access to commercial quantum capabilities, but
this access is not widespread, nor is it focused on hybrid quantum-classical software development.
See Adrian Cho, After Years of Avoidance, Department of Energy Joins Quest to Develop Quantum
avoidance-department-energy-joins-quest-develop-quantum-computers; Air Force Research
Laboratory to Join IBM Q Network as First DOD-led IBM Q Hub, Wright-Patterson AFB (Aug. 2, 2019),
https://www.wpafb.af.mil/News/Article-Display/Article/1924271/air-force-research-laboratory-to-join-
ibm-q-network-as-first-dod-led-ibm-q-hub/.
previous-reports/. In the FY 2021 NDAA, Congress took the first step toward implementing the
Commission’s First Quarter recommendations by creating a task force to develop a road map for the
NAIRR. See Pub. L. 116-283, William M. (Mac) Thornberry National Defense Authorization Act for
Fiscal Year 2021, 134 Stat. 3388 (2021). In Chapter 11 of this report, the Commission urges Congress
to authorize and appropriate the funds necessary to carry out the task force’s road map immediately.
30 Dan Strumpf, U.S. vs. China in 5G: The Battle Isn’t Even Close, Wall Street Journal (Nov. 9, 2020),
https://www.wsj.com/articles/u-s-vs-china-in-5g-the-battle-isnt-even-close-11604959200.
31 Dana Deasy, Department of Defense Statement on Mid-Band Spectrum, U.S. Department of
Defense (Aug. 10, 2020), https://www.defense.gov/Newsroom/Speeches/Speech/Article/2307288/
department-of-defense-statement-on-mid-band-spectrum.
32 Milo Medin & Gilman Louie, The 5G Ecosystem: Risks & Opportunities for DoD, Defense Innovation
Board (April 3, 2019), https://media.defense.gov/2019/Apr/03/2002109302/-1/-1/0/DIB_5G_
STUDY_04.03.19.PDF.
fcc.gov/wireless/bureau-divisions/mobility-division/35-ghz-band/35-ghz-band-overview.
34 Edward Drocella, et al., Technical Feasibility of Sharing Federal Spectrum with Future Commercial
technical-feasibility-sharing-federal-spectrum-future-commercial-operations-3450-3550.
35 See Public Notice: Auction of Priority Access Licenses in the 3550-3650 MHZ Band Closes, FCC
(Sept. 2, 2020), https://docs.fcc.gov/public/attachments/DA-20-1009A1.pdf.
36 In April 2019, the Defense Innovation Board issued a report which argued that the status quo
of spectrum allocation is unsustainable and DoD must expand its sub-6GHz spectrum-sharing
operations to enable the United States to compete with China in 5G. Milo Medin & Gilman Louie,
media.defense.gov/2019/Apr/03/2002109302/-1/-1/0/DIB_5G_STUDY_04.03.19.PDF. For additional
information and views of the Commission on 5G, see First Quarter Recommendations, NSCAI at 54-57
(March 2020), https://www.nscai.gov/previous-reports/.
37 Recommendations to ensure U.S. leadership in biotechnology, quantum computing, and 5G
telecommunications can be found above. Recommendations pertaining to semiconductors and
advanced computer hardware are covered in Chapter 13 of this report.
DSBSS15.pdf.
39 China owns one-fifth of the global supply of industrial robots and sought to have 45% of its high-
end robots be produced domestically by the end of 2020. See Johnny Williamson, How Nations
themanufacturer.com/articles/how-nations-around-the-world-are-investing-in-robotics-research/.
40 Intelligent Systems Division, NIST (last accessed Feb. 2, 2021), https://www.nist.gov/el/intelligent-
systems-division-73500.
p
597
ASSOCIATED TECHNOLOGIES
41 For example, the U.S. Postal Service could scale its Autonomous Mobile Robot pilot program from
25 sorting facilities to all sorting facilities by 2025. Autonomous Mobile Robots and the Postal Service,
U.S. Postal Service Office of Inspector General (April 9, 2018), https://www.uspsoig.gov/sites/default/
files/document-library-files/2019/RARC-WP-18-006.pdf.
42 An Executive Order 13806 report identifies 10 manufacturing risk archetypes. See Assessing
and Strengthening the Manufacturing and Defense Industrial Base and Supply Chain Resiliency
oct/05/2002048904/-1/-1/1/assessing-and-strengthening-the-manufacturing-and%20defense-
industrial-base-and-supply-chain-resiliency.pdf.
43 Strategy for American Leadership in Advanced Manufacturing, National Science and Technology
Council (Oct. 2018), https://trumpwhitehouse.archives.gov/wp-content/uploads/2018/10/Advanced-
Manufacturing-Strategic-Plan-2018.pdf; Gregory Tassey, Rationales and Mechanisms for Revitalizing
documents/2017/05/09/manufacturing_strategy_paper_0.pdf.
44 Report to the President on Ensuring American Leadership in Advanced Manufacturing, President’s
Council of Advisors on Science and Technology (June 2011), https://obamawhitehouse.archives.
gov/sites/default/files/microsites/ostp/pcast-advanced-manufacturing-june2011.pdf; Advanced
Manufacturing: A Snapshot of Priority Technology Areas Across the Federal Government, National
Science and Technology Council (April 2016), https://www.manufacturing.gov/sites/default/
files/2018-01/nstc_sam_technology_areas_snapshot.pdf.
45 Report to the President on Ensuring American Leadership in Advanced Manufacturing, President’s
Council of Advisors on Science and Technology at 3 (June 2011), https://obamawhitehouse.archives.
gov/sites/default/files/microsites/ostp/pcast-advanced-manufacturing-june2011.pdf.
46 Advanced Manufacturing: Innovation Institutes Have Demonstrated Initial Accomplishments, but
Challenges Remain in Measuring Performance and Ensuring Sustainability, GAO-19-409 at 1 (May 23,
2019), https://www.gao.gov/assets/700/699310.pdf.
47 Audit of the DoD’s Use of Additive Manufacturing for Sustainment Parts, U.S. Department of
Defense Inspector General (Oct. 17, 2019), https://media.defense.gov/2019/Oct/21/2002197659/-1/-
1/1/DODIG-2020-003.PDF.
48 Mark Anderson, 3D Print Jobs Are More Accurate With Machine Learning, IEEE Spectrum (Feb. 19,
2020), https://spectrum.ieee.org/tech-talk/artificial-intelligence/machine-learning/3d-print-jobs-news-
accurate-machine-learning.
49 For instance, in August 2020, the Department of Defense printed the first metal part for a B-52 jet
engine. See Kyle Mizokami, The Old-School Engine That Powers the B-52 Gets a 3D-Printed Upgrade,
Popular Mechanics (Aug. 10, 2020), https://www.popularmechanics.com/military/aviation/a33535790/
air-force-3d-print-metal-part-turbofan-engine/.
50 Robert Rapier, Ten Countries That Dominate Fossil Fuel Production, Forbes (July 14, 2019), https://
www.forbes.com/sites/rrapier/2019/07/14/ten-countries-that-dominate-fossil-fuel-production; Country
org/Statistics/View-Data-by-Topic/Capacity-and-Generation/Country-Rankings.
oe/energy-storage.
energy.gov/sites/prod/files/2020/12/f81/Energy%20Storage%20Grand%20Challenge%20Roadmap.
pdf.
p
598
APPENDICES
Appendices
p
599
THE NATIONAL SECURITY COMMISSION ON ARTIFICIAL INTELLIGENCE
Appendix A: Technical Glossary
601
Appendix B: Acronyms Found in This Report
617
Appendix C: Key Considerations for the Responsible
633
Development and Fielding of Artificial Intelligence (Abridged)
663
Appendix D: Draft Legislative Language
729
Appendix E: Funding Recommendation Table
Appendix F: Commissioner Bios
741
Appendix G: Commission Staff and Contributors
749
p
600
APPENDIX A
Appendix A:
Technical Glossary
3D Chip Stacking: The process of building integrated circuits with both horizontal and
vertical interconnections between transistors. This brings elements of the chip physically
closer together, increasing density and allowing for greater performance (i.e., speed) at
lower power levels and at a smaller footprint than comparable two-dimensional devices,
which only feature horizontal interconnects.
Additive Manufacturing: A computer-controlled process in which successive layers of
material are deposited to create a part that matches a 3D design.
Adversarial Machine Learning: A broad collection of techniques used to exploit
vulnerabilities across the entire machine learning stack and lifecycle. Adversaries may
target the data sets, algorithms, or models that an ML system uses in order to deceive
and manipulate their calculations, steal data appearing in training sets, compromise their
operation, and render them ineffective.1 Adversarial AI may be used as a phrase that
broadens the considerations to attacks on AI systems, including approaches that are less
dependent on data and machine learning.
Agile: A philosophy and methodology used to describe the continuous, iterative process
to develop and deliver software and other digital technologies. User requirements and
feedback inform incremental development and delivery by developers.2
AI Assurance: The defensive science of protecting AI applications from attack or
malfunction.
AI Digital Ecosystem: A technology stack driving the development, testing, fielding, and
continuous update of AI-powered applications. The ecosystem is managed as a multi-
layer collection of shared AI essential building blocks (e.g., data, algorithms, tools, and
trained AI models) accessed through common interfaces.
AI Governance: The actions to ensure stakeholder needs, conditions, and options are
evaluated to determine balanced, agreed-upon enterprise objectives; setting direction
through prioritization and decision-making; and monitoring performance and compliance
against agreed-upon directions and objectives.3 AI governance may include policies on
the nature of AI applications developed and deployed versus those limited or withheld.
p
601
TECHNICAL GLOSSARY
AI Lifecycle: The steps for managing the lifespan of an AI system: 1) Specify the system’s
objective. 2) Build a model. 3) Test the AI system. 4) Deploy and maintain the AI system. 5)
Engage in a feedback loop with continuous training and updates.4
AI Stack: AI can be envisioned as a stack of interrelated elements: talent, data, hardware,
algorithms, applications, and integration.5
Algorithm: A series of step-by-step instructions or calculations to solve an instance of
a problem. There are fundamentally two ways that algorithms are implemented by AI:
explicit engineering of the algorithm (e.g., in symbolic reasoning and expert systems) or by
machine learning, where the algorithm is derived from data or feedback from interactions.
Anonymization: Also referred to as data de-identification, this is the process of removing
or replacing with synthetic values any identifiable information in data. This is intended to
make it impossible to derive insights on any specific individual in the data while remaining
useful for the intended use of the data.6 (See de-anonymization.)
Application Programming Interfaces (APIs): Programming tools for describing how one
program can access the functionality of another7 while hiding the implementation details
inside each program.
Application-Specific Integrated Circuit (ASIC): A chipset custom designed to perform a
particular task. ASICs could provide significant performance gains over generic chips but
are inflexible in their functions compared to central processing units.
Architecture: A set of values, constraints, guidance, and practices that support the
active evolution of the planning, designing, and construction of a system. The approach
evolves over time, while simultaneously supporting the needs of current customers. 8
Architecture can refer to sets of components in a computing system and their operational
interrelationships as well as other important configurations such as the architecture of a
neural network, which captures the patterns of connectivity within and between layers of
units in the network model.
Artificial General Intelligence (AGI): A phrase that has been used to capture the possibility
of developing more general AI capabilities, in distinction to the typically narrow capabilities
of AI systems that have been developed to date. Some use the term to refer to the prospect
of achieving more human-like intelligence, developing AI systems with the ability to perform
many of the intellectual tasks that humans are capable of doing, or developing systems
that might employ a wide range of skills across multiple domains of expertise.
Artificial Intelligence (AI): The ability of a computer system to solve problems and to perform
tasks that have traditionally required human intelligence to solve.
p
602
APPENDIX A
Auditability: A characteristic of an AI system in which its software and documentation
can be interrogated and yield information at each stage of the AI lifecycle to determine
compliance with policy, standards, or regulations.
Augmented Reality: Enhanced digital content, spanning visual, auditory, or tactile
information, overlaid onto the physical world.9
Authorization to Operate (ATO): The official management decision given by a senior
organizational official to authorize operation of an information system and to explicitly accept
the risk to organizational operations (including mission, functions, image, or reputation),
organizational assets, individuals, and other organizations based on the implementation of
an agreed-upon set of security controls.10
Automation Bias: An unjustified degree of reliance on automated systems or their outcomes.
Autonomous: A system with functions capable of operating without direct human control.
Biological Sensors (Biosensors): Devices used to detect the presence or concentration
of a biological analyte, such as a biomolecule, a biological structure, or a microorganism.
Biosensors consist of three parts: a component that recognizes the analyte and produces
a signal, a signal transducer, and a reader device.11
Biometric Technologies: Technologies that leverage physical or behavioral human
characteristics that can be used to digitally identify a person and grant access to systems,
devices, or data, such as face, voice, and gait recognition.12
Black Box: The nature of some AI techniques whereby the inferential operations are complex,
hidden, or otherwise opaque to their developers and end users in terms of providing an
understanding of how classifications, recommendations, or actions are generated and
what overall performance will be.
Carbon Nanotubes: Nano-scale structures that can be used to make transistors and could
potentially replace silicon transistors in the future. Compared to existing silicon transistors,
carbon nanotube transistors are both capable of being shrunk to a smaller size and more
amenable to being stacked in three dimensions (see 3D chip stacking).
Cloud Computing: The act of running software within information technology environments
that abstract, pool, and share scalable resources across a network.13
Cloud Infrastructure: The components needed for cloud computing, which include
hardware, abstracted resources, storage, and network resources.14
p
603