Anthropic warns IPO investors of existential risks from AI technology

Anthropic has put potential investors on notice: its AI models carry risks that could threaten humanity's long-term survival. The disclosure appears prominently in the company's IPO prospectus, shared with select partners ahead of what is expected to be one of the largest public offerings ever.

Existential warnings and safety concerns#

The roughly trillion-dollar AI startup devoted substantial portions of its regulatory filing to cataloging potential dangers. Beyond extinction-level risks, Anthropic flagged that its technology could be weaponized for manipulation, blackmail, and unpredictable behavior as models grow more capable. These aren't hypothetical concerns buried in fine print—they're front and center in disclosures meant to inform investment decisions.

Business concentration and financial realities#

Alongside philosophical warnings, Anthropic revealed more traditional business risks. The company faces significant customer dependency, with a single quarter of its revenue flowing from just two clients. This concentration leaves Anthropic exposed if major customers shift strategies or reduce spending.

The finances behind building cutting-edge AI are equally stark. Anthropic plans to commit $518 billion to cloud, computing, and infrastructure over coming years—underscoring the capital intensity of competing at the frontier of AI development.

Record valuation and autumn listing#

Despite these headwinds, backers remain bullish. The company expects to list on Nasdaq this autumn at a valuation topping $2 trillion—more than double its May funding round and surpassing even Elon Musk's SpaceX. Investors point to Anthropic's growth trajectory as justification for the premium valuation, even as they grapple with existential risk disclosures and concentrated revenue streams.