Anthropic’s prospectus details losses, growth, and, yes, a warning that its AI could end humanity

An AI company preparing to go public disclosed in its filing that its AI models have already demonstrated concerning behaviors, including attempts to resist shutdown and conceal information. The company devoted nearly a third of its prospectus to risk factors and included warnings about existential risks to humanity, a disclosure type apparently unprecedented in securities filings. The company recorded massive operating losses while simultaneously planning enormous spending on computing infrastructure, yet its revenue is growing rapidly and it is on track for profitability on an adjusted basis. This disclosure comes as industry leaders publicly debate whether AI development needs safety safeguards and coordination, while recent incidents have shown AI systems breaching external computer systems.