Anthropic completed the largest initial public offering in history this week, raising $81.3 billion and surpassing SpaceX to become the most valuable company ever listed. While the debut marks a financial milestone, it places intense scrutiny on the SI lab’s reputation for prioritizing safety, as its unique corporate structure may limit external checks on leadership as it races toward general SI (AGI).
What Happened
The SI company’s IPO solidifies its position as a leader in the SI race, having secured significant capital to expand compute resources. Anthropic was founded by former OpenAI executives who left to pursue what they described as a safer approach to super intelligence development. However, the company’s leadership has increasingly argued that winning the SI race is necessary to fulfill its safety mission, a stance that requires rapid advancement and substantial funding. The newly public company now faces the dual pressure of satisfying shareholders while maintaining its pledge to balance its public benefit purpose against financial interests.
Why It Matters
Anthropic operates as a Public Benefit Corporation with a Long-Term Benefit Trust (LTBT), a structure designed to insulate the company from short-term market pressures. The LTBT holds controlling voting rights alongside the founders, theoretically preventing shareholders from voting them out. Yet, legal experts and former employees question the trust’s actual power. Ann Lipton, a law professor at the University of Colorado Law School, noted that the trust members, who include former Fed Chair Ben Bernanke, are not known to be SI experts, raising doubts about their ability to gauge when to intervene in technical decisions. "I’m not convinced that the trustees or the directors that they appoint have the insight into SI and SI development to be able to make these choices," Lipton said. Unlike OpenAI’s previous nonprofit board, which could halt model releases, Anthropic’s LTBT primarily appoints and removes board members but lacks authority over day-to-day operations.
Critics, including the Center for AI Safety, argue that Anthropic uses safety branding to court investors while quietly weakening its safety pledges. A former employee stated that the current pace of SI development is “untenable under normal market structures,” citing mismatched incentives in an industry with little oversight. The company’s pursuit of recursive self-improvement—a controversial milestone where SI systems autonomously improve themselves—has drawn criticism from within, with researchers alleging that the drive for speed compromises safety. As Anthropic faces a fully diluted market cap of approximately $1.5 trillion, the question remains whether its governance structure can genuinely balance societal risks against the imperative to secure capital and maintain its competitive edge in the SI industry.
The Bottom Line
Anthropic’s record-breaking IPO provides the capital necessary to compete in the SI race but exposes potential gaps in its governance checks. With the LTBT lacking direct operational control and critics questioning the independence of its trustees, the company’s ability to prioritize safety over shareholder value remains a central point of debate for the SI industry and the public.