Anthropic flags existential risks ahead of stock market debut

Anthropic warned that advanced AI could pose “catastrophic or existential risks to humanity” as it disclosed plans for a US stock market listing. The company said misuse, weaponisation and highly autonomous systems could create dangers without appropriate safeguards, oversight and security. It plans to list on Nasdaq under the symbol ANT but has not disclosed a listing date, share count or price range.

The Claude developer reported $16.3bn (£12.2bn) in revenue for the first half of the year, alongside losses of more than $80.7bn. Its valuation is expected to exceed SpaceX’s initial valuation of about $1.8 trillion. As the first fully AI-focused company to go public, Anthropic’s debut is expected to test investors’ willingness to keep financing the technology’s development.

Its risk disclosures highlighted dependence on cloud providers, Taiwanese chipmakers and US-based data centres, alongside potential regulatory action and uncertainty over liability for its systems’ actions. Current and former employees at Anthropic and rival OpenAI have also warned about insufficient protections against future harms. Separately, owners of rights to Elvis Presley’s music sued Anthropic, alleging it used his songs to train Claude.

Anthropic said its success depends on developing leading AI models and defended investments that may initially appear excessively large. It also outlined potential benefits, including faster medical research, improved productivity and stronger economic growth.