Support federal action to protect Americans from potentially catastrophic AI risks, including dangerous biological and cyber capabilities and unpredictable advanced systems.
Commitment kind: outcome. Promised action: Support government action to address catastrophic AI risks.. Promised outcome: Americans are protected from catastrophic AI harms involving biological, cyber, or uncontrollable AI capabilities.. Deadline: unknown; not assumed expired. As U.S. Senator from New Jersey during a 2026 Senate floor speech calling for a special AI session of Congress.
Occurrences
There are three challenges before us right now. First, we must protect against potentially catastrophic harms. Serious people, including researchers and leaders inside the companies developing the most advanced AI, are warning about dangerous biological or cyber capabilities—and the emergence of systems that behave in ways humans cannot reliably predict or control. We do not know whether the most catastrophic scenarios will ever occur. But when credible experts warn of consequences this grave, responsible government does not panic and it does not dismiss them. It gets to work.
Booker said government must act to protect against catastrophic AI harms such as biological, cyber, and loss-of-control risks.
unverified · Source version 63000 · locator unknown
"dangerous biological or cyber capabilities—and the emergence of systems that behave in ways humans cannot reliably predict or control"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): Within the lookback window, Booker explicitly urged federal attention to catastrophic AI risks matching the claim, but the source documents advocacy for a special session rather than enacted protections.
unverified · Source version 63000 · locator unknown
"Rigorous pre-deployment testing of advanced models"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): Booker listed concrete AI safety interventions he said he had advocated, including testing, audits, incident reporting, supply-chain security, frontier-model-weight security, and limits when capabilities outrun safeguards.
unverified · Source version 63009 · locator unknown
"Mr. Booker (for himself, Mr. Rounds, and Mr. Heinrich) introduced the following bill"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): GovInfo shows Booker introduced S. 3809, an AI research bill whose covered challenge categories include cybersecurity and AI safety-related cross-cutting challenges; the bill remained at the introduced/referred stage in this text record.
unverified · Source version 63011 · locator unknown
"Mr. Wyden (for himself, Ms. Warren, Mr. Booker"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): GovInfo records Booker as a cosponsor of S. 2164, a federal AI accountability bill directing FTC impact assessments for automated decision systems; the record shows referral to Senate Commerce, not enactment.
unverified · Source version 63013 · locator unknown
"Cory Booker (D-N.J.) reintroduced" "safety, reliability, and trustworthiness of AI"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): A Senate office release says Booker joined the CREATE AI Act reintroduction, framing the NAIRR proposal as advancing safer, more reliable, trustworthy AI research; this is a legislative effort, not final delivery.
unverified · Source version 63015 · locator unknown
"keep the most advanced artificial intelligence (AI) models out of the hands of non-state actors"
Unverified model/legacy excerpt; not proof. AI summary (separate from source): Booker joined Senate Foreign Relations Democrats backing U.S.-China AI safety talks and strict export controls, relevant to dangerous AI capabilities and security risks, but the statement does not establish enacted protections.