Anthropic Warns of 'Catastrophic Risk' as It Readies Blockbuster Stock Market Listing

Anthropic has put a warning in the opening pages of its stock market paperwork: advanced artificial intelligence, it says, could pose "catastrophic or existential risks to humanity." The language appears in a filing made public on Monday, ahead of the company's planned listing in the United States, and it comes from a firm whose own product, the Claude chatbot, is the thing being sold to investors.

The company says it will trade on the Nasdaq under the symbol ANT. It will be the first fully AI-focused company to go public, which makes the listing a test case for how markets treat the technology. Its valuation is expected to exceed that of SpaceX, which held the biggest listing on record this summer at an initial valuation of about $1.8 trillion. Anthropic has not yet said how many shares it will release, at what price range, or on what date.

The financial figures are stark. Anthropic reported $16.3bn (£12.2bn) in revenue for the first half of the year, against losses of more than $80.7bn. The company acknowledged that it will often make forward investments "that may seem irresponsibly large at first glance," and offered a motto for its approach: "bold but sane," winning trust through a record of successful bets. It also said its fortunes depend largely on building the most advanced AI models, and that failing to do so could hurt its business.

The risk section runs to about 90 pages. On safety, the company said concerns had been raised by policymakers, regulators and researchers, and by people inside its own organisation. Advanced systems could pose catastrophic or existential risks, it said, if not developed, deployed and operated with appropriate safeguards, oversight and security, particularly if they were misused, replicated, weaponised or allowed to operate with a high degree of autonomy.

The filing also lists more practical vulnerabilities. Anthropic relies on a cluster of cloud computing providers, on Taiwanese chipmakers already facing huge demand, and on data centres concentrated in the US. It flagged possible regulatory action over harms linked to its models, citing concerns about use by children and by people at risk of suicide, and said it is uncertain where and how it could be held liable for what its systems do.

Alongside the warnings, the company made its case for the technology. It called AI "one of the most consequential technologies in human history," and described a future of "millions of virtual agents with the ability to revolutionise every area of human endeavour." It said AI could "profoundly accelerate the rate of progress in biology and medicine," potentially finding cures for diseases including cancer and Alzheimer's, and could produce "sustained acceleration in economic growth" through productivity gains.

The backdrop is crowded. Anthropic was founded five years ago by chief executive Dario Amodei, his sister Daniela Amodei and five others, and announced its intention to list in June. Rival OpenAI quickly followed, and according to a person familiar with its plans, its listing is now expected early next year. Both announcements were followed by a wave of cyber-security and other technological breaches involving their AI agents, bots built to operate somewhat autonomously. Current and former staff at both companies have also warned that the technology is being developed without the ability to protect humanity from possible future harms caused by AI tools. On the day of the filing, the owners of the rights to Elvis Presley's music sued Anthropic, accusing it of using hundreds of his songs to train Claude.

Analysts framed the filing as expected rather than shocking. Matt Orton, chief market strategist at Raymond James Investment, said investors anticipated both rapid growth and rapid losses, and that the details should give the market more visibility, since it needs to know where revenue growth is coming from. Peter Andersen, founder of Andersen Capital Management, told Reuters the plan was "a significant milestone in the development of artificial intelligence," and that the debut's outcome would be "fairly indicative of the sentiment" among investors about continuing to pour money into AI. Pricing, share count and a date are the next things to watch, followed by OpenAI's listing early next year.