As Anthropic prepares for a potential blockbuster stock market debut, prospective public investors face an unusual reality: an external group of trustees controls the majority of the AI company’s board. The planned initial public offering, which could value the maker of Claude at as much as EUR 2 trillion, is drawing fresh scrutiny to the lab’s experimental governance model.
At the center of that model sits the Long-Term Benefit Trust (LTBT), a small group of advisers established to protect Anthropic’s stated mission of developing AI for the long-term benefit of humanity, even as commercial pressures grow. The trust holds no equity in the company, yet it wields significant influence. The San Francisco-based firm intends to keep the LTBT’s role intact after going public.
How the Long-Term Benefit Trust Works
As the controlling governance mechanism of Anthropic’s public benefit corporation, the LTBT can appoint or dismiss the majority of the company’s board. It has already selected four of Anthropic’s seven directors, including Netflix co-founder Reed Hastings and Novartis chief executive Vas Narasimhan.
The trust currently has three members out of a possible five. It is chaired by Neil Buddy Shah, chief executive of the Clinton Health Access Initiative, and includes former Federal Reserve chair Ben Bernanke and Richard Fontaine, CEO of the Center for a New American Security. Mariano-Florentino “Tino” Cuéllar, a former California Supreme Court justice, departed after a few months to become Anthropic’s chief global affairs officer.
According to people close to the company’s plans, Anthropic hopes to create a template for future AI governance. The goal is for the LTBT to serve as an industry blueprint, comparable to GAAP accounting principles, which began as voluntary standards proposed by private companies. Bernanke’s appointment in July signaled the intent to build a more institutional body.
Access, Authority, and an Untested Structure
The trustees must receive advance notice of major company actions, including the launch of new AI models. They meet weekly among themselves and convene with Anthropic’s leadership as often as every other week. They also attend regular board meetings and discuss significant issues directly with the company’s founders.
Those discussions have included the company’s Mythos cybersecurity model, where trustees encouraged a limited rollout through the Glasswing Project, as well as a dispute with the U.S. government over automated weapons.
Despite that access and authority, the trust has functioned largely in an advisory capacity. It has not attempted to set firm limits or force a major trade-off between profit and purpose, according to a person familiar with its operations. As a result, the structure has yet to face the kind of conflict that would reveal whether it can restrain Anthropic’s leadership when commercial and societal goals diverge.
The arrangement creates a “built-in conflict,” according to Jesse Fried, a Harvard Law School professor and corporate governance expert. The company “raises funds from profit-seeking investors, then lets self-appointed individuals decide how much profit to sacrifice for the firm’s mission—ensuring its AI benefits humanity,” he wrote.
Source
Image: arstechnica.com
