Anthropic warns investors its AI could pose 'catastrophic or existential risks to humanity' as it files for IPO

From the Newshound archive · first published 30 September 2026.

Anthropic's IPO prospectus contains what may be an unprecedented investor warning: the Claude-maker's own technology could produce "catastrophic or existential risks to humanity," including AI attempts to "resist shutdown," "conceal or manipulate information" and conduct "resembling blackmail." The filing also discloses that the company posted a GAAP net loss of $41.97 billion last year, even as revenue surged 1,088% to $4.59 billion. Anthropic safety researcher Evan Hubinger put the odds of AI killing humans within the next decade at above 10%.

Anthropic's $518 billion AI buildout is tied to non-cancellable commitments, the filing shows.

The prospectus also lays bare the company's heavy reliance on Big Tech partners, raising questions about its independence as a public company.