Anthropic, founded in 2021 with the mission to ensure long-term human control and benefit from AI, is approaching its IPO with an unusually complex governance structure that limits public shareholders' control. Unlike typical public companies, Anthropic is set up as a public benefit corporation, allowing its board to prioritize AI safety and public mission alongside profitability without being liable to shareholder lawsuits. The board currently includes co-founders Daniela Amodei as president and her brother Dario Amodei as CEO, as well as notable figures like Netflix co-founder Reed Hastings and former Microsoft CFO Christopher Liddell.

A key factor in Anthropic’s unique setup is the Long-Term Benefit Trust (LTBT), an independent body wielding substantial influence by controlling a special stock class that can appoint and remove a majority of the board’s directors—referred to as Class T directors. The trust is chaired by Neil Shah, a global health expert, and features other influential members such as former Federal Reserve chair Ben Bernanke and national security expert Richard Fontaine. This LTBT is charged with maintaining focus on Anthropic’s mission to benefit humanity long-term, further insulating the company’s decisions from typical investor pressures.

Beyond the LTBT, Anthropic’s seven founders collectively hold a special voting stock class granting them 50.1% of shareholder voting power, enabling them to elect three board members. This arrangement effectively concentrates control with the founders initially, but includes provisions to phase out this privileged setup ten years after the IPO or if fewer than three founders remain. At that point, the LTBT will continue its governance role until a major vote or decision ends its oversight, at which time common shareholders might gain more influence—though only far in the future.

The board itself has controls to check the LTBT’s power, including employee-backed votes to approve certain board nominations or removals. This multi-layered structure echoes past tensions in AI startups like OpenAI, where board and employee interests can clash. In Anthropic’s case, investors on public markets will have limited direct influence initially, with key decisions resting in the hands of founders, the LTBT, and the board, reflecting the company’s commitment to AI safety above ordinary shareholder control.