Anthropic IPO to Preserve External Trust's Control Over Board
nashnova research
Anthropic is pushing toward an IPO at a valuation of up to $2 trillion, but an external trust holding zero equity controls the majority of board appointments — a governance structure never stress-tested, now set to become public-market investors' defining risk.
What exactly is this trust?
Anthropic created the Long-Term Benefit Trust (LTBT) to control board personnel — it appoints four of seven directors, giving it a majority.
The critical detail: trust members hold no Anthropic equity whatsoever. There are currently three members, expandable to five.
The three members are Neil Buddy Shah (chair, CEO of the Clinton Health Access Initiative), former Fed Chair Ben Bernanke, and Richard Fontaine (CEO of the Center for a New American Security).
This means → a group of outsiders with no financial stake in the company holds effective control over its board — an extremely unusual setup for a tech IPO.
Can the trust actually constrain the company?
The trust has the right to advance notice of major corporate actions, including new AI model launches, and meets with management every two weeks.
But sources say the trust has so far operated as an advisory body — it has never attempted to draw a red line or force a trade-off between profit and mission.
In plain terms = the trust has the power but has never actually used it — no one knows whether it can withstand real commercial pressure.
Harvard Law professor Jesse Fried put it bluntly: the company raises money from profit-seeking investors while letting self-appointed individuals decide how much profit to sacrifice — "a deep and potentially unmanageable tension hard-coded into the company's corporate DNA."
What does the OpenAI precedent tell us?
In November 2023, OpenAI's board tried to fire CEO Sam Altman but failed after losing the confidence of investors and employees. Several directors were replaced, and the company underwent a broader structural overhaul.
This reflects a hard lesson: when governance structures collide with commercial reality, the pushback from capital and talent can overwhelm the board's formal authority.
Whether Anthropic's trust structure can avoid the same outcome remains unanswered — because it has never been truly tested.
What is Anthropic trying to build with this structure?
Sources say Anthropic wants the LTBT to become a governance template for the AI industry — similar to U.S. Generally Accepted Accounting Principles (GAAP), which started as a voluntary corporate initiative before becoming an industry standard.
Bernanke joined the trust in July this year, a move seen as a signal that Anthropic is building the body into a more institutionally authoritative entity.
University of Pennsylvania Law professor Elizabeth Pollman noted that fierce competition in AI — both geopolitical and corporate — makes balancing profit and mission increasingly difficult.
In plain terms = Anthropic is betting that a trust with no equity and no battle record can keep constraining a $2 trillion company after it goes public — whether that bet pays off is the single most important question for investors.
市场有风险，内容仅供研究参考，不构成投资建议。