Anthropic Warns of AI Risks Including Shutdown Resistance in IPO
Anthropic, an AI startup behind Claude AI, warned investors in its IPO prospectus about catastrophic risks from advanced AI. The company highlighted potential autonomous behaviors like resisting shutdown, hiding information, or blackmail-like actions. CEO Dario Amodei urged caution in AI development, while researchers estimate over 10% chance AI could kill humans within 10 years. About 80 pages of the 261-page prospectus detail these risks.
HT World·3h ago


