Warren Sends Anthropic 11-Page Demand on Existential Risks Before IPO

Senator Elizabeth Warren sent an 11-page letter to Anthropic CEO Dario Amodei demanding answers about existential AI risks before the company proceeds with its IPO, citing Anthropic's own SEC preliminary registration statement, which includes 90 pages of risk factors. Warren requested answers by November 3 and warned that AI risks demand greater scrutiny before they become entangled with public capital markets. Her concerns span both public safety and potential systemic risks to financial markets from Anthropic's size and share-price volatility. Warren is the top Democrat on the Senate Banking, Housing and Urban Affairs Committee. Separately, researcher Geoffrey Irving and journalist Sigal Samuel have publicly argued that Anthropic and other leading AI labs should stop unilaterally rather than wait for imminent danger. Irving contended that AI labs are operating on 'slapdash empirics,' patching environments without a coherent long-term safety strategy, and that waiting until a model is directly dangerous assumes governments can coordinate internationally on a very short timeline. Daniel Kokotajlo criticized Dario Amodei for calling for pacing without committing to unilateral action.