The company behind the Claude family of models has spent its short life raising capital privately, much of it from the cloud providers that also supply its computing power. Its registration statement arrived at a moment when enthusiasm for AI-linked stocks had been wavering. Stock markets around the world moved higher on Tuesday, a rebound for AI-exposed shares that Bloomberg attributed to the prospect of the listing after a difficult stretch for the sector. A listing would give public investors their first direct stake in one of the handful of foundation model companies working at the frontier of the technology.
The document leaves its headline terms blank, as first drafts typically do. The price range and share count will be filled in closer to the sale, after regulators review the filing and the company meets prospective buyers.
Where the prospectus is specific is on how the company will be governed once it is public. Anthropic is incorporated as a public benefit corporation, obliged to weigh a stated purpose of developing AI responsibly against shareholder returns, and an independent Long-Term Benefit Trust holds a special share class that elects a majority of the board while a founder-controlled entity holds another class carrying just over half the voting power on other matters, according to the filing. Founder-weighted share structures are familiar in large technology listings, but routing board control through a trust of outside trustees is unusual for a company of this size.
The sale would test investor appetite at a scale rarely attempted. Its chief executive has described advanced AI as a potential existential threat to humanity, yet the company is heading for a debut as soon as next month that would beat rival OpenAI to the public market and could surpass the record SpaceX set with its June listing, the AP reported. The timetable now depends on the Securities and Exchange Commission's review and the amendments that usually follow.