Anthropic was founded in 2021 with the stated purpose of ensuring humanity can control and benefit from artificial intelligence. As the company moves toward an initial public offering (IPO), ordinary investors who buy its stock on public markets will face a governance structure that substantially limits their influence.
Public benefit company status and mission protection
Anthropic is organized as a public benefit company, a legal form that allows a for-profit corporation to weigh a public mission — here, promoting AI safety — alongside shareholder value. That status gives the board legal cover to make safety-driven decisions that are not solely focused on maximizing investor returns.
Current board composition
The board is chaired by Anthropic president Daniela Amodei. Other members include CEO Dario Amodei (her brother), Reed Hastings (Netflix cofounder), Christopher Liddell (former Microsoft and General Motors CFO), Vas Narasimhan (Novartis CEO), and Yasmin Razavi (partner at Spark Capital). There is one vacant seat on the board.
Long Term Benefit Trust (LTBT) and Class T directors
A significant portion of board control is allocated to a Long Term Benefit Trust (LTBT), an independent panel intended to keep the company focused on its mission to benefit humanity. The trust owns a class of shares that enables it to elect and remove a majority of the company’s board — these are the Class T directors.
The LTBT is chaired by Neil Shah, formerly managing director at GiveWell. Its members include former Federal Reserve chair Ben Bernanke, national security expert Richard Fontaine, and Harvard physicist Christopher Stubbs. The LTBT currently has one vacancy, and existing members will select the replacement.
Founders’ special voting class
Anthropic’s seven founders jointly hold a special class of stock that gives them combined voting power equal to 50.1% of shareholder votes if a majority of the founders act together. That voting bloc can elect three members of the board of directors. If a majority of the founders does not agree, their special shares vote proportionally with common stockholders.
Board checks and employee voting
There is an additional restraint on the LTBT’s power: when the LTBT nominates a “Specified” one of its four directors, or removes such a director without cause, the board’s nominating committee can refuse to approve the action and instead refer the decision to the company’s employees for a vote. This mechanism formalizes a route for employee influence over board composition.
Practical implications and precedents
This layered, baroque governance makes outcomes difficult to predict. A notable precedent is OpenAI’s recent governance crisis, when the board briefly fired CEO Sam Altman and reversed course after employees rallied; that episode led to a new governance arrangement. Anthropic’s model appears to institutionalize a role for employee power alongside founder and trust controls.
Depending on how unified the founders remain, common shareholders could end up with more direct influence than initially expected, or conversely see major decisions shaped by the LTBT or by employee votes if the board takes that route.
Sunset clauses and end conditions
The founders’ special control is not necessarily permanent: ten years after the IPO the founders may choose to end their control, and it will automatically end if fewer than three founders remain. After that point, the LTBT continues until either 85% of common shareholders vote to terminate it or the trust opts to step down with board approval.
Only when neither the founders’ special class nor the LTBT exists will common shareholders obtain full control — a transition the company’s structure delays until specific conditions are met.
