Anthropic's Secret Board: Who Really Controls the AI Company?

Anthropic, one of the leading AI companies, is getting ready to sell shares to the public for the first time. But before that happens, it's worth understanding who actually controls the company. It turns out that a special trust, called the Long Term Benefit Trust (LTBT), owns just one special share of Anthropic. That single share gives the trust significant power over the company's board of directors and major decisions, like whether to transfer a powerful AI model.

The board has seven seats, and the LTBT gets to appoint some of them. The trust is governed by a private agreement that isn't public, so we don't know exactly how its trustees are chosen or how it makes decisions. Anthropic says trustees serve one-year terms and elect their own replacements. The trust also has to be notified before certain actions, like transferring a majority of shares or a capable AI model, so it can step in if it disagrees.

This lack of transparency is concerning because Anthropic's stated goal is to develop AI responsibly for humanity's benefit. If we don't know who controls the company, we can't hold it accountable. The company should publish the trust agreement and related voting agreements, especially as it prepares for an IPO. Until then, the public is left in the dark about who really calls the shots at one of the most important AI firms in the world.