Tim Fist is the Director of Emerging Technology Policy at the Institute for Progress (IFP), a Washington, D.C. think tank. His published work concentrates on the physical and regulatory infrastructure of AI development — data-centre and power build-out in the United States, semiconductor export controls, and the policy response to increasingly automated AI research and development (Source: ifp.org).
Background
Fist is an adjunct senior fellow at the Center for a New American Security, where he was previously a fellow. Before joining CNAS he was head of strategy at Fathom Radiant, an AI hardware company, and a senior machine learning engineer at Eliiza, an AI software company. His writing has appeared in Foreign Policy and ChinaTalk, and his research has been cited in Wired, Reuters, Inside AI Policy, The Wire China, Breaking Defense and the Wall Street Journal (Source: ifp.org).
Compute and infrastructure
Fist's longest-running line of work treats domestic compute capacity as a policy variable rather than a market outcome. The Compute in America series, begun in June 2024 with Arnab Datta and Brian Potter, examines what would be required to build the next generation of AI infrastructure inside the United States; its second part, "How to Build the Future of AI in the United States" (October 2024), and the February 2025 policy playbook on gigawatt-scale cluster construction extend the argument toward permitting, transmission and siting (Source: ifp.org). A related June 2024 paper argues for a non-profit foundation attached to NIST as a mechanism for accelerating the agency's emerging-technology work.
Export controls and China
Fist has co-authored a sequence of IFP assessments on whether specific NVIDIA accelerators should be exportable to China: "Should the US Sell Blackwell Chips to China?" on the B30A (October 2025, with Georgia Adamson, Saif Khan and Tao Burga), "Should the US Sell Hopper Chips to China?" on the H200 and H100 (December 2025, with Khan, Burga and Adamson), and "When Do More AI Chips for China Mean Fewer for the United States?" (March 2026, with Adamson), which assesses whether H200 production for China would divert supply from US customers given AI-chip supply constraints (Source: ifp.org). An earlier April 2025 paper, "The H20 Problem: Inference, Supercomputers, and US Export Control Gaps," argued that gaps in the controls then in force would permit the sale of 1.3 million AI chips to China. He testified to the US House Committee on Science, Space, and Technology in April 2025 on the lessons the United States should draw from DeepSeek, and co-authored "America's AI Exports Program" (March 2026) on the strategic goals and implementation of US AI export promotion. See Compute Governance and US-China AI Competition: Different Races, Different Metrics.
Automated AI research and pacing
With Saif M. Khan and five IFP colleagues — Tao Burga, Arthur Tellis, Ben Schifman, Jonah Weinbaum and Olivia Scharfman — Fist published "How Should the US Prepare for Increasingly Automated AI R&D?" on August 6, 2026, setting out 23 policy recommendations described as low-regret How Should the US Prepare for Increasingly Automated AI R&D? (IFP, August 2026) (Source: ifp.org).
Fist and Khan restated the argument in a guest post dated August 9, 2026, responding to the July 2026 "Pacing the Frontier" statement. They name three risks that a capacity to pace automated AI research would address — offence-dominant capability uplift, loss of control, and power concentration — propose that pacing be specified by identifying which automated research activities pose severe risks and by reallocating resources toward diffusion and safety research, list seven government preparatory actions, and cite the AI Data Center Moratorium Act as the counterproductive default they expect if preparation does not happen Should we \"pace\" AI self-improvement? (Fist and Khan, August 2026). See Recursive Self-Improvement (RSI).
Provenance note: The August 6, 2026 IFP report and the August 9 guest post are queued for foundational ingest; citations here should upgrade to the primary texts once those pages exist.
Other work
Fist has contributed to IFP's agenda-setting output on AI policy more broadly: "An Action Plan for American Leadership in AI" (March 2025), twelve recommendations submitted to the Office of Science and Technology Policy's AI Action Plan process; an April 2025 analysis of the public recommendations that process received; "Catalyzing a Golden Age" (June 2025), a response to the OSTP request for information on the 2025 National AI R&D Strategic Plan; and "Preparing for Launch" (August 2025), the introduction to IFP's Launch Sequence project, for which he later co-authored a rolling request for proposals (Source: ifp.org). He is one of the named authors of "Open Problems in Technical AI Governance". Outside AI, he co-authored an October 2025 paper on a coordinated federal effort to secure the US rare-earth supply chain.
Relationships
- affiliated-with: Institute for Progress (IFP), Center for a New American Security (CNAS)
- co-author: Saif M. Khan
- supports: Recursive Self-Improvement (RSI) (pacing of automated AI R&D), Compute Governance
- authored: Open Problems in Technical AI Governance (named co-author)
- related: Pacing the Frontier (statement from employees of frontier AI companies, July 2026), Artificial Intelligence Data Center Moratorium Act, US-China AI Competition: Different Races, Different Metrics, DeepSeek