Anthropic's $2 Trillion IPO Puts Experimental Governance Trust Under Public Scrutiny

Anthropic's planned initial public offering — which could value the Claude maker at as much as $2 trillion — is drawing intense scrutiny to an unusual governance structure that places a small group of external trustees in control of the company's board, raising questions about how that arrangement will hold up under the demands of public markets.

At the center of the structure is the Long-Term Benefit Trust, or LTBT, a body created to safeguard Anthropic's stated mission of developing artificial intelligence for the long-term benefit of humanity. The trust holds no equity in the company but wields significant authority: it has the right to appoint or dismiss the majority of Anthropic's seven-member board, and has already selected four of those directors.

Among those four are Reed Hastings, co-founder of Netflix, and Vas Narasimhan, chief executive of Novartis. The trust itself currently has three members, chaired by Neil Buddy Shah, chief executive of the Clinton Health Access Initiative, alongside former Federal Reserve chair Ben Bernanke and Richard Fontaine, CEO of the Center for a New American Security.

Bernanke's appointment in July was seen by people close to Anthropic's plans as a deliberate move to build a more institutional body — one that could serve as an industry blueprint for AI governance, similar to how GAAP accounting principles emerged as voluntary standards before becoming widely adopted norms, according to those people.

The trustees receive advance notice of major company actions, including the launch of new AI models, and meet weekly among themselves. They also convene with Anthropic's leadership as frequently as every other week, attend regular board meetings, and hold discussions with the company's founders on significant issues. Those have included the rollout of Anthropic's Mythos cybersecurity model — where trustees encouraged a limited release through the Glasswing Project — and the company's dispute with the U.S. government over automated weapons, according to a person close to the talks.

Despite that access, the trust has largely operated in an advisory capacity and has not yet forced a meaningful trade-off between profit and safety, said a person familiar with its workings — meaning the structure has not yet faced the kind of conflict that would test whether it can genuinely constrain Anthropic's leadership.

That dynamic has drawn skepticism from legal scholars. Jesse Fried, a professor at Harvard Law School and corporate governance expert, wrote in a July paper that the arrangement creates a "built-in conflict." The company "raises funds from profit-seeking investors, then lets self-appointed individuals decide how much profit to sacrifice for the firm's mission," he wrote, adding that "a deep and potentially unmanageable tension is thus hard-wired into the firms' corporate DNA."

Elizabeth Pollman, a law professor at the University of Pennsylvania, echoed those concerns. "It's nearly impossible to perfectly contract for all possible circumstances that could arise when managing competing interests within a firm," she said, pointing to geopolitical and corporate competition in AI as factors that complicate the balancing act. "Will this governance structure work in the way intended, serving dual or more interests over time? That's the real challenge."

One structural safeguard distinguishes Anthropic's model from OpenAI's — the trust can be dissolved with the support of 85 percent of shareholders' voting power, a provision that functions as a kind of "kill switch." OpenAI's board became a cautionary reference point in November 2023, when it attempted to fire CEO Sam Altman, lost the confidence of investors and employees, and ultimately saw most of its directors replaced.

A person close to Anthropic said its private investors backed the company with full knowledge of its governance structure, and that several cited the emphasis on safety as part of their investment rationale. Still, at least one venture capitalist who has backed Anthropic offered a blunter read: "There was a judgment made by investors that capitalism would win in the end."

As Anthropic moves toward a public listing, its trustees will face a broader and potentially less patient investor base — one that will be watching closely to see whether a governance model designed for an idealistic mission can survive the pressures of a public company balance sheet.

Disclaimer