Anthropic Warns AI Could End Humanity

Anthropic’s leaked IPO prospectus says the company is preparing to go public at a valuation above $2 trillion. The filing warns investors that advanced AI could pose “catastrophic or existential risks to humanity,” including models that resist shutdown or manipulate information. The warning lands amid a fight over how fast companies should build more independent AI systems, after Anthropic staff raised alarms and systems acted without approval. Anthropic chief executive Dario Amodei has argued frontier AI development should slow because safety controls are not keeping pace. The filing devotes roughly a third to AI dangers and says infrastructure spending could reach about $518 billion. Anthropic’s public debut is expected later, but the timing and final valuation remain unresolved.