Anthropic’s Post‑IPO Governance: Who Actually Controls the AI Frontier?

Governance Architecture of a Public Benefit Company

Anthropic, founded in 2021 to ensure humanity can control AI and reap long‑term benefits, is preparing for an initial public offering while deliberately limiting the influence of ordinary shareholders. The company’s structure combines a public benefit corporation status, a Long‑Term Benefit Trust (LTBT), and a special founder share class to concentrate decision‑making power.

The Long‑Term Benefit Trust’s Dominance

As a public benefit company, Anthropic can prioritize AI safety alongside profit without fear of shareholder lawsuits. The LTBT, chaired by global health expert Neil Shah, holds a separate class of stock that allows it to elect and remove a majority of the board—known as Class T directors. Its current roster includes former Federal Reserve chair Ben Bernanke, national security expert Richard Fontaine, and Harvard physicist Christopher Stubbs. One seat on the LTBT remains vacant, to be filled by existing members. This trust’s control means that common shareholders cannot sway most board appointments.

Founder Voting Power and Board Seats

The seven founders collectively own a special class of stock that grants them 50.1 % of voting power, enabling them to appoint three board members. Only if the founders fail to reach a majority do their shares vote proportionally with common stock. The present board, chaired by president Daniela Amodei, includes her brother CEO Dario Amodei, Netflix co‑founder Reed Hastings, former Microsoft and GM CFO Christopher Liddell, Novartis CEO Vas Narasimhan, and Spark Capital partner Yasmin Razavi. One board seat is currently open.

Employee Check on the LTBT

A built‑in safeguard allows the board’s nominating committee to reject a “Specified” LTBT director nomination or removal without cause, sending the decision to an employee vote. This mechanism mirrors the employee‑driven backlash that forced OpenAI’s board to reinstate CEO Sam Altman after his brief ouster—a case often cited as a precedent for Anthropic’s governance design.

What Happens When Founder Control Ends

Founder control is not permanent. Ten years after the IPO the founders may terminate their majority, or control automatically lapses if fewer than three founders remain. At that point the LTBT persists unless 85 % of common shareholders vote to dissolve it, or the trust steps down with board approval. Only when both the founder share class and the LTBT no longer exist will ordinary shareholders gain full voting control—assuming, of course, the company survives any potential AI‑related catastrophe.

Implications for Shareholders and Employees

The layered structure means common shareholders will have little sway over board composition and strategic direction for the foreseeable future. Employees, however, retain a decisive voice in contested director appointments, giving them leverage over the LTBT’s choices. As the company moves toward public markets, the tension between mission‑driven governance and shareholder expectations will shape Anthropic’s trajectory.