Who controls powerful AI? Anthropic has a plan for long after its leaders depart

San Francisco-based AI lab Anthropic has devised a complex governance plan that aims to preserve its mission and balance investor interests, even as its leaders transition out of the company. With an expected valuation of $1.5 trillion, Anthropic operates as a public benefit corporation under Delaware law, necessitating a delicate equilibrium between profit and promoting societal well-being.

The company's seven co-founders hold 50.1% voting power on specific matters, but an independent trust controls the majority of board seats. This structure allows the co-founders, equipped to steward Anthropic's mission, to maintain influence in elections and shareholder proposals. Upon most co-founders leaving and their supervoting stock gone, the board will split into three groups with three-year terms, akin to the staggered US Senate elections.

This structure diverges from current corporate governance norms, where major US companies typically elect all directors annually. Anthropic's long-term benefit trust, chaired by health leader Neil "Buddy" Shah, controls Class T stock, which can elect a majority of the board. The trust's rules demand adherence to Anthropic's public benefit purpose, even if the stock is relinquished.

The company's existing safety and risk committee, including Novartis CEO Vas Narasimhan and former White House official Chris Liddell, will remain integral to governance. Billionaire Netflix co-founder Reed Hastings serves as the lead independent director. Despite these measures, co-founders, including the Amodei siblings, retain the core role in governance, with the Founder LLC entity holding the most significant influence in board elections.