Anthropic files for IPO while warning of catastrophic AI risks

Anthropic has filed for an initial public offering and laid out plans for a Nasdaq listing, pairing its push toward public markets with warnings about the potential dangers of its technology. Its prospectus says the systems could pose “catastrophic or existential risks to humanity” if developers do not use appropriate safeguards.

The company reported first-half revenue of US$16.3 billion and losses of US$80.7 billion. The Sydney Morning Herald likewise highlighted the prospect of losses in the billions and Anthropic’s warning about existential harm.

Thisismoney reported the same first-half financial figures and said two unnamed customers generated 23 per cent of revenue, a concentration detail not included in the other accounts. The expected scale of the listing varies by measure and source. Thisismoney said markets expect the IPO could raise as much as $100 billion at a valuation of around $2 trillion.

Insurance Business described Anthropic as seeking a Nasdaq valuation expected to exceed US$1.8 trillion. Those are expectations, not completed fundraising or a final market valuation. Quartz’s headline also reported the IPO filing and Nasdaq plan, identifying “ANT” as the planned listing symbol. The filing also raises questions about legal responsibility for AI systems.

Insurance Business reported that Anthropic disclosed uncertainty over where and how it might be held liable for actions taken by its systems. The company’s warning about safeguards and its liability disclosure put the risks of developing and deploying the technology alongside the financial case for a public listing.

Insurance Business framed the prospectus as relevant to AI liability and insurance governance, and said the Claude system is embedded in financial-services, insurance and professional-services operations across Canada.

It argued that Canada’s proposed Artificial Intelligence and Data Act would not address the issue quickly enough, saying Bill C-27 remained before Parliament and had not passed. These policy and industry points are that outlet’s emphasis; the other reports focus on the filing, financial disclosures and risks.

The prospectus’s language on possible harm is notably stark, while the company’s strategy is described in the same account as “bold but sane.” The filing therefore presents both a growth proposition and a warning about the consequences of insufficient safeguards, without resolving how liability for system actions would be determined.

The next concrete questions are whether the offering proceeds to a Nasdaq debut and what valuation investors ultimately place on Anthropic. The reported figures remain forecasts, while the company’s first-half revenue and losses are disclosed results; the sources do not give a final offering price or listing date.