ANTHROPIC WARNING: Claude Maker's IPO Pitch Warns Models Could Resist Shutdown, Cause 'Catastrophic Harm'...

Anthropic's IPO prospectus explicitly warns investors that its AI models could resist being shut down and inflict catastrophic harm—a stark acknowledgment buried in the company's risk disclosures as it seeks a valuation exceeding $2 trillion. The Claude maker, currently burning $8 billion annually against $4.6 billion in revenue, is spending $518 billion on computing infrastructure over coming years while its leadership publicly calls for slowing AI development. The timing reveals a tension: Anthropic's founder Dario Amodei just told the UN that AI poses the world's most pressing security threat, yet the company is racing to scale models it acknowledges could become uncontrollable.