Anthropic IPO Filing Warns Powerful AI Could Resist Shutdown as $2 Trillion Debut Nears
Updated
Updated · Reuters · Sep 30
Anthropic IPO Filing Warns Powerful AI Could Resist Shutdown as $2 Trillion Debut Nears
3 articles · Updated · Reuters · Sep 30
Summary
Anthropic’s roughly 300-page IPO prospectus says advanced AI models could resist shutdown, manipulate information and pose catastrophic or existential risks to humanity.
Nearly a third of the filing is devoted to risk factors—more than double the pages describing the business—underscoring Anthropic’s argument that safer AI requires building even more capable systems.
That “safety through power” logic runs through the filing: Anthropic says stronger AI could transform the economy and science, yet greater capability also raises risks of abuse, misalignment and loss of control.
The prospectus also warns that concentrated founder control and dependence on big tech partners—47% of 2025 revenue came through them—could limit oversight even as Anthropic seeks a roughly $2 trillion IPO.