Anthropic plans $2 trillion IPO, putting its unique Long-Term Benefit Trust in the public spotlight
The Claude maker is preparing a massive public debut that will test whether public markets can tolerate a non-equity trust holding absolute board veto power.
Original cover art, generated for this story. THE VISSION does not republish third-party press imagery.
- Anthropic is planning an initial public offering targeting a historic $2 trillion valuation, a move that will put its unconventional governance model to the test.
- The company is governed by a Long-Term Benefit Trust holding no equity but possessing the sole authority to appoint or dismiss four out of seven board directors.
- The three-member trust, chaired by Neil Buddy Shah, met weekly to dictate safety-focused launch limits on the company's recent Mythos model via the Glasswing Project.
- Harvard Law Professor Jesse Fried warns that the structure hard-wires a deep and potentially unmanageable tension between safety and shareholder returns.
Anthropic is preparing an initial public offering (IPO) that could value the artificial intelligence startup at $2 trillion, a monumental debut that represents a major test for public-market tolerance of unconventional corporate governance, according to reporting from Ars Technica and the Financial Times. Unlike standard corporations where shareholders elect directors, Anthropic is controlled by an independent body known as the Long-Term Benefit Trust (LTBT), which holds the absolute authority to appoint and dismiss a majority of the company's board of directors.
The trust consists of independent trustees who hold zero equity or financial stake in Anthropic, insulating them from short-term financial pressures to prioritize its public benefit mission. Currently chaired by Neil Buddy Shah, CEO of the Clinton Health Access Initiative, the trust also includes former Federal Reserve Chairman Ben Bernanke and Richard Fontaine, CEO of the Center for a New American Security. A former fourth trustee, Mariano-Florentino 'Tino' Cuéllar, recently stepped down to become Anthropic’s Chief Global Affairs Officer, leaving two vacancies on the trust.
The trustees meet weekly to monitor company operations and model launches. In a recent exercise of their authority, the trust intervened in the rollout of Anthropic's Mythos cybersecurity model. Under an initiative dubbed the Glasswing Project, the trustees mandated a highly restricted, defensive-only release of Mythos, demonstrating their active role in model safety oversight. To address investor risk, Anthropic's structure includes a 'kill switch' where 85% of shareholder voting power can dismiss the trustees, a key difference designed to prevent the rigid gridlock seen during OpenAI's 2023 board crisis.
However, legal and financial experts are skeptical of how public markets will handle the trust. Jesse Fried, a Harvard Law School professor, noted that 'a deep and potentially unmanageable tension is thus hard-wired into the firms' corporate DNA,' as the trust's safety mandate may directly conflict with shareholders' demands for financial returns. An anonymous venture capitalist echoed the sentiment, telling reporters there was 'a judgment made by investors that capitalism would win in the end,' suggesting a belief that public-market pressure will eventually force the trust to align with profit-seeking motives.
Anthropic's IPO is a landmark event for corporate law and AI safety. If successful, it proves that a multi-trillion-dollar technology giant can go public while keeping its steering wheel firmly in the hands of independent trustees focused on public benefit over shareholder value. If it fails or faces intense legal challenges from public shareholders, it will signal that Wall Street cannot tolerate non-equity safety mechanisms, forcing future AI labs to choose between access to public capital and safety-first corporate architectures.