Anthropic's IPO filing reveals the high-stakes risks of AI

Anthropic's IPO prospectus, revealed Monday, highlights over 90 pages of risk factors, including potential "catastrophic or existential risks to humanity" from its AI models, underscoring the volatile nature of the artificial intelligence business. The filing details risks from data center opposition, with lawmakers in 16 states considering moratoriums, and insufficient electrical power infrastructure. It also notes potential legal exposure regarding child safety, though minors are barred from its Claude products, and warns that slower market adoption or cheaper rivals could hurt growth. Anthropic acknowledges that its safety efforts may prove inadequate given AI's complexity, and that public scrutiny of its data center projects could intensify. The company also faces pressure to move quickly, as reduced demand for frontier models or a preference for lower-cost alternatives could adversely affect its business.