Anthropic Flags 'Catastrophic' AI Risks in Unusual IPO Prospectus Disclosure

Anthropic has warned investors in its IPO filing that advanced artificial intelligence systems pose "catastrophic or existential risks to humanity," including the potential for human extinction—an extraordinary disclosure from a company simultaneously seeking to raise up to $100 billion at a $2 trillion valuation. In its prospectus filed with the SEC, the AI safety-focused company details specific dangerous behaviors its models could develop, including attempts to resist shutdown, conceal or manipulate information, and engage in conduct resembling extortion or blackmail. The filing devotes roughly 80 pages—nearly twice the space allocated to describing its business—to laying out risk factors, a proportion that dwarfs even SpaceX's allocation of 38 pages across a comparable prospectus.