2024-01707. AI in Global Development Playbook

AGENCY:

United States Agency for International Development (USAID)

ACTION:

Notice; request for information.

SUMMARY:

The United States Agency for International Development and the U.S. Department of State, in coordination with the National Institute of Standards and Technology (NIST), seek information to assist in carrying out responsibilities under Executive Order 14110 on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence issued on October 30, 2023. Specifically, the E.O. directs USAID and the State Department to publish an AI in Global Development Playbook that incorporates NIST's AI Risk Management Framework's principles, guidelines, and best practices into the social, technical, economic, governance, human rights, and security conditions of contexts beyond United States borders.

DATES:

Comments containing information in response to this notice must be received on or before March 1, 2024. Submissions received after that date may not be considered.

ADDRESSES:

Comments may be submitted by any of the following methods:

Sent as an attachment to [e-mail] in any of the following unlocked formats: HTML; ASCII; Word; RTF; Unicode, or .pdf.

Written comments may be submitted by mail to: USAID, IPI/ITR/T, Rm. 2.12–213, RRB, 1300 Pennsylvania Avenue NW, Washington, DC 20004.

Response to this RFI is voluntary. Submissions must not exceed 10 pages (when printed) in 12-point or larger font, with a page number provided on each page. Please include your name, organization's name (if any), and cite “AI in Global Development Playbook” in all correspondence.

Comments containing references, studies, research, and other empirical data that are not widely published should include copies of the referenced materials. All comments and submissions, including attachments and other supporting materials, will become part of the public record and subject to public disclosure.

USAID will not accept comments accompanied by a request that part or all of the material be treated confidentially because of its business proprietary nature or for any other reason. Therefore, do not submit confidential business information or otherwise sensitive, protected, or personal information, such as account numbers, Social Security numbers, or names of other individuals.

FOR FURTHER INFORMATION CONTACT:

For questions about this RFI contact: Andrew Merluzzi, [e-mail] or 1–802–558–5397.

Accessible Format: USAID will make the RFI available in alternate formats, such as Braille or large print, upon request by persons with disabilities.

SUPPLEMENTARY INFORMATION:

To promote safe, responsible, and rights-affirming development and deployment of AI abroad, the Executive Order on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence directs:

“The Secretary of State and the Administrator of the United States Agency for International Development, in coordination with the Secretary of Commerce, acting through the director of NIST, [to] publish an AI in Global Development Playbook that incorporates the AI Risk Management Framework's principles, guidelines, and best practices into the social, technical, economic, governance, human rights, and security conditions of contexts beyond United States borders. As part of this work, the Secretary of State and the Administrator of the United States Agency for International Development shall draw on lessons learned from programmatic uses of AI in global development.

USAID and the State Department are seeking information to assist in carrying out this action.

With the right enabling environment, ecosystem of market actors, and investments, AI technologies can foster greater efficiency and accelerated development results across a variety of sectors and contexts, whether in agriculture, health, education, energy, etc. Addressing the risks presented by AI technologies is essential to fully harnessing their benefits. Understanding these risks across a range of geographic and cultural contexts requires the expertise of local communities, the private sector, civil society, governments, and other stakeholders.

The AI in Global Development Playbook aims to characterize the risks and opportunities of AI in Global Majority countries (sometimes referred to as low- and middle-income countries, developing countries, or the “Global South”), and will provide guidance for various stakeholders—organizations building, deploying, and using AI; private sector; governments; and others—to address those risks and leverage opportunities to drive AI applications for sustainable development. This RFI is an attempt to collect various research products, experiences, and perspectives that will inform the Playbook and speak to the unique risks and benefits of the use of AI technologies in Global Majority countries, including concrete examples of successes, hurdles, and roadblocks.

AI ecosystems are the stakeholders, systems, and an enabling environment that empower people and communities to build and use AI tools responsibly, as well as to respond to the use of AI technologies in their contexts. While no two AI ecosystems are identical, there are broad characteristics that many ecosystems share or upon which they differ. Feedback on this RFI will help illuminate the most impactful ecosystem factors and inform how best to navigate those factors to advance a responsible approach to AI.

In considering information for submission, respondents are encouraged to review resources that USAID, State Department, and NIST have developed or coordinated with partners to develop in the past:

• USAID Digital Ecosystem Framework
• USAID AI Action Plan
• Reflecting the Past, Shaping the Future: Making AI Work for International Development
• NIST AI Risk Management Framework
• OECD Working Party on AI Governance
• Global Partnership on AI

• OECD Recommendation on AI

• Hiroshima Process Code of Conduct for Organizations Developing Advanced AI Systems

1. Questions for the AI in Global Development Playbook

USAID and State Department are interested in receiving information pertinent to any or all of the topics described below. Respondents may provide information on one or more of the topics in this RFI and may elect not to address every topic.

Please answer based on your experience, the positions of your organization, or research you have encountered or conducted. Where possible, please cite the source of your information or note when personal views are expressed.

Information that is specific and actionable is of special interest. Copyright protections of materials, if any, should be clearly noted. USAID and the State Department are especially interested in the perspectives of those living and/or working in Global Majority countries, though responses are welcome from anyone.

The Opportunities, Risks, and Barriers of AI

1. What are the most important barriers in Global Majority countries to achieving a future where AI tools are designed and deployed in a responsible way to address the UN Sustainable Development Goals (SDGs) and support humanitarian assistance. How would you address these barriers?

2. What applications of AI or AI technologies are most promising for advancing the SDGs and supporting humanitarian assistance? How can these applications be advanced responsibly? Are there any sectors that are particularly well suited to applications of AI? Are there potential limitations or trade-offs that should be considered when applying AI in these contexts?

3. Relatedly, what are the most risky or harmful applications of AI in Global Majority countries? Why? Can their risks or harms be mitigated, and if so, how?

4. How are commercially available AI tools currently helpful in addressing the SDGs and supporting humanitarian assistance? Where do they fall short or lead to harm? What steps have or should be taken to mitigate such harms?

5. How do AI's potential benefits and risks differ for specific groups (particularly disadvantaged or marginalized groups) and between geographic and cultural contexts?

The Enabling Environment for Responsible AI

6. How should data for AI systems be collected, used, stored, managed, and owned to further the SDGs and support humanitarian goals? Which aspects of data management are unique or particularly salient for AI? How should the objective of ensuring sufficient data accessibility for AI training be reconciled with other objectives, such as ensuring privacy protections, in different contexts?

7. What kind of AI-related financial and resource investments should actors in Global Majority countries prioritize to achieve the SDGs and support humanitarian assistance? What kinds of financing and resourcing is most needed to catalyze responsible AI development?

8. How should computational resources (“compute”) to build or deploy AI systems be managed in Global Majority countries? How could compute be more accessible, affordable, and reliable? How should hardware and infrastructure to support the deployment of AI systems be managed and governed?

9. What are the barriers to building the AI workforce in Global Majority contexts, including for tasks beyond technical development of AI systems? What kinds of skills or experience are most needed in these contexts? Where can people gain these skills and experiences?

10. What other AI-enabling infrastructure or resources are needed to advance responsible AI development and use?

AI Policy, Protections, and Public Participation

11. Are there existing AI principles, tools, or best practices that you think are particularly helpful in advancing AI for development in a risk-aware manner? If they are only partly helpful, where do they fall short?

12. What kinds of AI-related policies do you think are most promising (or harmful) in Global Majority contexts? Why? Who might these policies benefit, and who might they harm? How might existing policies be reshaped for improved outcomes?

13. How might AI affect broader labor-market dynamics in your context? Are there some skills for which it increases demand, and others for which it decreases demand?

14. How might AI affect competition dynamics in your context? Do these effects vary by economic sector?

15. How should the public be informed about AI risks and harms in your context, and engaged on AI governance issues? What efforts around community engagement seem promising? What communities should be engaged who are not part of existing discussions?

16. What are the best ways to improve inclusivity and stakeholder representation in AI design, deployment, governance, or policymaking in the context of global development (at the global, regional, and local levels)?

17. What are best practices for ensuring human rights are respected and protected in the development, deployment, and use of AI in the context of a risk-based approach to AI governance? Are there mechanisms, processes, and capacity in place to hold actors accountable for harms resulting from AI systems in your context? What should be done to create and operationalize those accountability mechanisms, and ensure their sustainability?

18. Please list any other organizations you think should be consulted as the AI in Global Development Playbook is developed (please note it may not be possible to consult with every organization).

Authority: Executive Order 14110 of Oct. 30, 2023.

Signing Authority

This document of the United States Agency for International Development was signed on January 24, 2024, by Andrew Merluzzi, Emerging Technology Advisor, USAID. That document with the original signature and date is maintained by USAID. For administrative purposes only, and in compliance with requirements of the Office of the Federal Register, the undersigned USAID Federal Register Liaison Officer has been authorized to sign and submit the document in electronic format for publication, as an official document of the USAID. This administrative process in no way alters the legal effect of this document on publication in the Federal Register .

Signed in Washington, DC, on January 24, 2024.

Andrew Merluzzi,

Emerging Technology Advisor, USAID.