Anthropic's 335-page stock filing promises an age of miracles and warns of existential risk. Anyone thinking of buying AI stocks should know what it leaves out.
Anthropic, the AI company behind Claude, has filed a 335-page document to sell its stock to the public. It pairs big promises with blunt warnings. Its founders talk about bringing forth "an age of miracles." Page 46 says AI could "pose catastrophic or existential risks to humanity." The filing was made Monday, and it is unusual even by tech standards.
The document is called an S-1. It is the long report a company files before it sells shares to the public for the first time. Ninety-four of its pages are risk factors, the section where a company lists what could go wrong. A letter from the founders asks buyers "not to think of your purchase in purely transactional terms" and to see it as a chance to be "part of something greater."
The filing is candid. It says AI models could "behave in ways that no user or developer intended," leading to incidents so large the whole insurance industry could not cover them. It also admits the company may make spending bets that "may seem irresponsibly large." CEO Dario Amodei has said there is an up to 25 percent chance that further AI development could lead to human extinction.
The founders call this "radical honesty" and part of an "anti-sell" approach. They say it is dangerous when people buy a stock based on expectations the company cannot meet. Anthropic will also go public as a public benefit corporation. That kind of company must weigh a social mission alongside shareholder interests.
Filings were not always like this. Jay R. Ritter, a University of Florida business professor, noted that Apple's 1980 prospectus ran 47 pages with zero risk factors. Facebook's filing, he said, was a turning point at 200 pages with 22 pages of risk factors. His complaint is that the bloat makes it harder to find what matters.
Some key numbers are absent. Third quarter results, which would help show whether the company makes money, are not included. Anthropic reported explosive revenue growth in the second quarter. But that quarter was also the height of what the article calls tokenmaxxing, which could have driven a surge in sales. Some corporate executives have warned that AI costs are surging out of control.
Anthropic's filing is candid about risk but light on recent profit data, so read it carefully before treating any AI stock as more than a long-term bet.
World models, AI that learns physics from video, are being built into home robots and driverless cars, and your own videos may help train them.
Read more: AI that learns how the physical world works is coming to robots and carsA new Alibaba-led test of real coding requests found the best AI agent fully finished just 26.4 percent of tasks, so human review still matters.
Read more: AI coding helpers fully finish only about 1 in 4 real tasksA University of Chicago law professor asked Claude to fix page numbers. It deleted his syllabus and a year of notes instead. Dropbox saved him, not the AI.
Read more: An AI assistant deleted a law professor's year of lecture notes
