U.S. Budget Cuts Target Agencies Tasked With Stopping AI-Enabled Bioterrorism
A CSIS analysis finds the FY2026 budget proposal guts the biosecurity and AI evaluation capacity of NIST and CAISI at the same moment the administration's own AI Action Plan expands their mandate.
The Trump administration's fiscal year 2026 budget proposal would strip hundreds of millions of dollars from the federal agencies it has simultaneously assigned to guard against artificial-intelligence-assisted bioterrorism, according to a policy analysis published this month by the Center for Strategic and International Studies.
The CSIS paper, authored by Georgia Adamson and Gregory C. Allen of the Wadhwani AI Center, argues that large language models and biological design tools are approaching a threshold where they could provide meaningful assistance to actors attempting to develop pathogens. The report states, in language that should be read as a moderate-confidence assessment rather than a confirmed capability finding, that commercial AI tools are already "on the cusp" of helping novices develop or acquire biological weapons by providing step-by-step guidance. The underlying concern is dual-use ambiguity that the authors decline to resolve falsely: the same models useful for drug discovery and pandemic response can lower informational barriers to harm.
The administration's July 2025 AI Action Plan named the National Institute of Standards and Technology and its subordinate Center for AI Standards and Innovation as primary government contacts for biosecurity testing of frontier AI models. That mandate expanded on a similar tasking in the Biden-era national security memorandum on AI. According to the CSIS analysis, the Action Plan also added new responsibilities, including monitoring foreign frontier AI development for national security implications.
Yet the FY2026 budget proposal, as reviewed in the CSIS paper and corroborated by public budget documents, would cut NIST funding by roughly $325 million, a nearly 30 percent reduction from its FY2025 level, with the justification citing the agency's support for what the administration characterized as a "radical climate agenda." The proposal also calls for eliminating close to 500 full-time NIST employees. CAISI, for its part, currently operates on approximately $15 million annually against an estimated need of $84 million to fulfill its AI Action Plan taskings, according to a separate analysis from the Institute for Progress.
The broader biosecurity budget picture is starker. The CSIS analysis notes that the same proposal threatens cuts of $18 billion to NIH, $3.6 billion to CDC, and $240 million to the Administration for Strategic Preparedness and Response. Those are proposed figures, not enacted law, and Congress retains appropriations authority. But the directional pressure is consistent across agencies.
The institutional hollowing-out extends to the White House itself. According to reporting by STAT News reviewed for this piece, Gerald Parker resigned from the National Security Council's Biosecurity and Pandemic Response directorate after roughly six months, having never been formally appointed to head the White House Office of Pandemic Preparedness and Response Policy. As the AEI-published account by co-author Britt Lampert documents, OPPR had 20 staff before the presidential transition and now has none; the NSC biosecurity directorate is left with a single part-time employee. Congress established OPPR in 2022 with bipartisan support specifically to ensure continuity of pandemic coordination across administrations.
The structural tension the CSIS paper identifies is not subtle. The administration is formally expanding CAISI's scope under its own AI Action Plan while proposing to shrink the budget and headcount required to execute that scope. Whether Congress acts to close that gap, or whether CAISI absorbs the worst cuts while other NIST offices bear the reduction, remains unclear. The CSIS authors assess the answer to that question as consequential: nucleic acid synthesis screening and frontier-model biosecurity evaluation are the two chokepoints they identify as most tractable for AI-era policy, and both require the government capacity that is now under pressure.
For now, the United States is moving in two directions at once: formally designating NIST and CAISI as its primary technical resources for AI-biosecurity evaluation, and proposing to defund and depopulate them at the same time. That is not a risk management posture. It is a gap.
Sources cited:
- CSIS Wadhwani AI Center - "Opportunities to Strengthen U.S. Biosecurity from AI-Enabled Bioterrorism" (https://www.csis.org/analysis/opportunities-strengthen-us-biosecurity-ai-enabled-bioterrorism-what-policymakers-should)
- STAT News - "White House pandemic office left leaderless after Parker exit" (https://www.statnews.com/2025/07/30/white-house-pandemic-preparedness-office-leaderless-unprepared/)
- The Hill / AEI - "White House pandemic team reduced to one part-time staff" (https://thehill.com/opinion/healthcare/5473546-white-house-pandemic-preparedness-collapse/)
- Institute for Progress - "What Will It Cost for the US to Be Ready for the Next Big AI Breakthrough?" (https://ifp.org/funding-for-caisi/)
- NIST - Center for AI Standards and Innovation (CAISI) (https://www.nist.gov/caisi)
This release was originally distributed via ETL Newswire. Visit CSIS Wadhwani AI Center - "Opportunities to Strengthen U.S. Biosecurity from AI-Enabled Bioterrorism" for the full story, related releases, and contact information.
Visit CSIS Wadhwani AI Center - "Opportunities to Strengthen U.S. Biosecurity from AI-Enabled Bioterrorism" →