Role: 46th President of the United States (2021–2025) Party: Democratic
Joe Biden served as the 46th President of the United States from 2021 to 2025. His presidency coincided with the public release of ChatGPT and the emergence of frontier AI as a federal policy question. Working with National Security Adviser Jake Sullivan and a cross-agency team, his administration assembled the federal AI-safety architecture of 2023–2024, anchored by Executive Order 14110, the creation of the US AI Safety Institute within NIST, and Bureau of Industry and Security chip export controls on China. Most of the executive-branch components of that architecture were rolled back or reoriented after January 2025.
Executive Order 14110
In October 2023 Biden signed Executive Order 14110, "Safe, Secure, and Trustworthy AI," the most comprehensive US executive-branch AI action issued to that date. Its principal elements included Defense Production Act reporting requirements, under which frontier developers above a compute threshold were to report training runs, red-team results, and safety evaluations to the federal government; operationalization of the NIST AI Risk Management Framework, including NIST AI 600-1 on generative AI; the seeding of the US AI Safety Institute; and agency implementation deadlines under which dozens of agencies issued AI guidance pursuant to EO 14110 authorities.
EO 14110 was rescinded by President Trump on January 20, 2025, and replaced by Executive Order 14365 — Ensuring a National Policy Framework for AI and America's AI Action Plan.
Two Office of Management and Budget memoranda, OMB M-24-10 and M-24-18, operationalized federal-agency AI governance under EO 14110 authorities. These remained partially in force despite the rescission of the parent order.
International AI diplomacy
Biden's administration participated in and shaped the Bletchley Declaration (UK, November 2023), the Seoul Frontier AI Safety Commitments (May 2024), and the Hiroshima AI Code of Conduct (G7, 2023). These processes produced the international AI Safety Institute network linking the UK AISI, US AISI, and counterparts in Japan, Singapore, the EU, and others. The Trump administration retained elements of the US AISI but stepped back from the multilateral safety posture, including by declining to sign the Paris AI Action Summit Declaration (2025).
Export controls
Under Biden, the Bureau of Industry and Security issued the October 2022 and October 2023 advanced-chip export control packages and the AI Diffusion Rule. Jake Sullivan coordinated these measures at the National Security Council level.
Policy positioning
Biden's frame treated frontier AI as a dual-use technology requiring both competition with China, through export controls and the CHIPS Act, and risk management, through the US AISI, Defense Production Act reporting, and the NIST framework. The administration's stated position was that US leadership in AI is compatible with, and depends on, credible safety institutions. The 2025 Trump rescission removed much of the risk-management component while retaining and expanding the competition component.
Sources
- EO 14110 and source summary
- BIS Framework for AI Diffusion — Interim Final Rule (Jan 13, 2025) and source summary
- The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), Frontier AI Safety Commitments (Seoul, 2024), G7 Hiroshima Code of Conduct for Advanced AI (2023)
- US AI Safety Institute (NIST AISI)
- US AISI Strategic Vision
Relationships
- supersedes: Executive Order 14110 — Safe, Secure, and Trustworthy AI (signed by Biden) was later superseded by Executive Order 14365 — Ensuring a National Policy Framework for AI (signed by Trump).
- contradicts: Donald Trump on AI regulation — the 2025 rescission of EO 14110 is the clearest break.
- related: Jake Sullivan — NSC lead on AI/compute geopolitics.
- related: US AI Safety Institute (NIST AISI), UK AI Safety Institute (AI Security Institute) — Biden administration was central to creating the AISI network.
- related: Bureau of Industry and Security (BIS) — Biden-era BIS export-control regime.
- instance-of: AI Safety Cases and Frameworks — EO 14110 was the principal US federal expression.