Anthropic's $2 Trillion IPO Dream Faces New Challenge: NYC Official Wants It Paused

New York City Comptroller Mark Levine is urging Anthropic to pause its planned initial public offering (IPO) until the artificial intelligence company addresses concerns about AI safety, corporate governance and the risks of rapidly advancing technology.

In a letter dated Oct. 14 to Anthropic CEO Dario Amodei, Levine argued that the company should prioritize safety measures and greater transparency before pursuing an IPO that could value it at approximately $2 trillion.

The letter raises questions about how public-market investors would assess Anthropic’s business as the company itself warns about the potential dangers of increasingly capable AI systems.

Levine pointed to growing concerns within Anthropic about the risks of increasingly powerful AI models. In a Sept. 12 essay, Amodei warned that AI capabilities could outpace society’s ability to control them and called for slowing the pace of development.

The comptroller also cited the Sept. 9 resignation of Anthropic researcher Jacob Coxon, who warned that some company insiders believe AI could pose an existential threat to humanity by the end of the decade.

80 Pages Of Risk Disclosures

Those concerns are also reflected in Anthropic’s draft IPO filing, which reportedly warns that advanced AI could pose "catastrophic or existential risks to humanity," Levine wrote.

Anthropic’s risk disclosures in its draft S-1 filing span about 80 pages, nearly twice as long as the section describing its business.

"These disclosures are not routine and indicate risks that are without precedent, let alone at a public offering of this scale," Levine wrote.

He also highlighted warnings from Anthropic’s leadership about the potential for AI agents to cause widespread damage. Amodei has warned that increasingly capable AI systems could enable coordinated cyberattacks and other disruptive activity without adequate safeguards.

"These concerns underscore the seriousness of the risks at issue. Even if these outcomes prove overstated, such dire warnings illustrate the broad economic risk that imprudent AI advances pose," Levine wrote.

Governance Concerns Could Complicate Public Listing

Beyond AI safety, Levine questioned whether Anthropic’s planned governance structure would provide public shareholders with sufficient accountability.

Anthropic is organized as a Delaware public benefit corporation and has a Long-Term Benefit Trust. Levine cited reports that the company intends to adopt a structure that could concentrate control among founders and trustees.

He argued that a potential gap between decision-making power and accountability is particularly concerning given the risks Anthropic has acknowledged.

Levine also criticized a recent agreement between Anthropic and five other companies with the White House, calling the "morally binding" accord insufficient to address the industry’s challenges.

Instead, he called on Anthropic to detail its planned audits and safeguards, explain what risks those controls would address and provide a timeline for implementation. He also asked the company to explain whether its proposed measures would support Amodei’s call to slow AI development.

"Prudence should not only be shown with respect to the pace of the frontier, it should be also shown with respect to the timing of an IPO," Levine wrote.

He asked Amodei for a written response and offered to discuss the concerns further.

Anthropic has not formally announced its IPO date, although sources previously said the Claude-maker plans to go public sometime in mid-November. Discussions are still underway, and the timing could change again, sources noted. 

Photo: Shutterstock