Loading page…
The Long-Term Benefit Trust can appoint or dismiss most directors while holding no shares. Its untested ability to balance mission and commercial pressure will become a public-market governance question.
Listen to this story
Anthropic’s prospective public listing would preserve a mission-governance structure in which the Long-Term Benefit Trust has selected four of seven directors and can appoint or remove a board majority. The trust currently has three members, including Ben Bernanke, and receives advance access to major decisions while holding no equity. Its power has not yet been tested in a direct profit-versus-mission clash.
Trustees influenced a limited Mythos cybersecurity-model rollout through Glasswing Project and discussed automated weapons policy, but have not vetoed a major commercial decision.
Shareholders holding 85% of voting power can remove trustees; public-listing terms could alter that threshold.
The trust has three current members—Neil Buddy Shah, Ben Bernanke, and Richard Fontaine—and can expand to five.
Anthropic plans to retain a trust with the power to appoint or dismiss a majority of its board after a prospective stock-market debut that could value the company at as much as $2 trillion. That gives potential public investors an unusual governance constraint to weigh alongside the AI lab’s commercial case.
Anthropic is a public benefit corporation, a corporate form intended to balance commercial success with public benefit. The Long-Term Benefit Trust holds no equity; Anthropic says trustees are independent of management and investors, do not share in profits, and are paid only for their time and service.
The trust currently has three members out of a possible five: Neil Buddy Shah, Ben Bernanke and Richard Fontaine. Trustees receive advance notice of major company actions, including model launches; meet weekly; attend board meetings; and regularly meet Anthropic’s leadership.
Those channels have put the trustees into discussions over the Mythos cybersecurity model, where they encouraged a limited rollout through the Glasswing Project, and Anthropic’s dispute with the U.S. government over automated weapons, according to a person close to the talks.
Despite that access and authority, the trust has largely acted as an adviser and has not forced Anthropic to choose between a major commercial objective and its mission. Whether it would exercise its formal leverage in such a conflict remains the central unresolved test.
Harvard Law School professor Jesse Fried argues the tension is built into the structure: profit-seeking investors fund the company while mission guardians may decide how much profit to sacrifice. University of Pennsylvania law professor Elizabeth Pollman said competing interests cannot be fully anticipated in a contract.
The shareholder removal mechanism is why the structure has been described as less risky than OpenAI’s, though still unproven; the required supermajority could change in a public listing. OpenAI offers a nearby caution: its board tried to fire chief executive Sam Altman in November 2023, then lost investor and employee confidence. Most directors were replaced, and the company restructured.
For investors, the question is not whether Anthropic has built a mission-oriented check. It is whether a small group without an economic stake can use its powers effectively when the company’s purpose and a consequential business decision diverge.
Loading discussion...
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.