Anthropic’s IPO prospectus warns that its advanced AI systems could pose “catastrophic or existential risks to humanity,” according to multiple reports citing the filing. The disclosures also highlight concerns that the models could behave in unexpected ways, including “self-preserving behaviours,” as capabilities and use cases expand.
Outlets describe the warning as unusual coming from a company seeking to profit from AI. Reuters-cited reporting says the company plans to underscore these risks to potential investors. NDTV and others note the filing devotes a substantial portion of its main prospectus to risk factors, while Business Line characterises the concern as growing with further development and broader deployment. Seeking Alpha and Free Malaysia Today similarly focus on the risk-heavy language used to caution investors.
