Anthropic went public on Thursday in the largest initial public offering ever, raising $81.3 billion and surpassing SpaceX for that record, according to the company. The debut sets a benchmark for other AI labs expected to follow it onto public markets. It also puts new scrutiny on Anthropic's claim that it develops AI more carefully than its competitors, a reputation the company has cultivated since its founders left executive research roles at OpenAI.

The structure Anthropic will operate under as a public company gives its leadership unusual latitude. Anthropic is registered as a Public Benefit Corporation, which requires its board to weigh its mission alongside shareholder returns. Its founders and its Long-Term Benefit Trust, introduced in 2023, hold controlling voting rights, meaning shareholders could not vote them out. A special class of stock, Class T, gives the trust's trustees the authority to select and remove some members of Anthropic's board.

The fine print narrows that authority. The trust picks four of Anthropic's seven board members, but one of those four effectively requires founder approval. The trust must also give 30 days' notice of nominations or removals to a nominating and corporate governance committee made up of co-founder Daniela Amodei, who chairs it, Netflix co-founder Reed Hastings, and Anthropic investor Yasmin Razavi. The trust's four listed members are former Fed Chair Ben Bernanke, foreign policy analyst Richard Fontaine, experimental physicist Christopher Stubbs, and former GiveWell executive Neil Buddy Shah, who chairs the trust.

Ann Lipton, a University of Colorado Law School professor who studies corporate governance, questioned whether the trust would use its board power to keep Anthropic on mission and whether trustees know enough about AI to judge when to slow a company plan. She noted the trustees' prominence but asked what Bernanke knows about AI. Tyler Johnston, founder of the nonprofit AI industry watchdog The Midas Project, said Anthropic could theoretically hire away a trustee with a large compensation offer that would require leaving the trust.

Hamza Chaudhry of the Future of Life Institute said neither the trust nor its trustees hold decision-making authority over day-to-day matters, and he does not expect the trust to serve as a key governance mechanism after the IPO. A former Anthropic employee who left before the offering said the founders hold essentially all real decision-making power and that trustees were chosen to avoid friction. Anthropic's Form S-1 indicates the trust's main duty is appointing and potentially removing certain directors, without the day-to-day or safety-related authority OpenAI's nonprofit board held. As a public company, Anthropic will face heightened pressure to raise capital and grow revenue, which the former employee said makes the current pace of AI development hard to sustain under normal market structures.

More company and startup news from TechManNews.