New York City Comptroller Mark Levine has asked Anthropic to pause its IPO process until the company puts stronger AI safety and governance measures in place. Bloomberg reported that Anthropic has since established a communication channel with the comptroller’s office, although Levine said he could not disclose what the two sides discussed.
The request lands as Anthropic prepares for what has been described as a potentially record-setting IPO. According to the report, Levine sent a letter on Oct. 14 Eastern Time to Anthropic chief executive Dario Amodei, asking the company to stop the offering process until it had addressed concerns about AI safety, governance, and the pace of research "fully and transparently."
Letter framed through the city pension funds’ fiduciary role
In the letter, Levine said he serves as an investment adviser, custodian, and trustee for New York City pension funds that manage more than $325 billion in assets. He described the system as one of the largest public retirement systems in the United States.
Levine wrote that Anthropic had recently acknowledged in public that frontier model capabilities may be advancing faster than society can understand or control them. He also pointed to Amodei’s Sept. 12 essay, We Must Pace the Frontier, which called for slowing the pace of improvements in AI model capabilities.
His argument was that if the company itself is issuing warnings of that scale, caution should also show up in the timing of its public listing.
Levine asked for details on audits, controls, and timing
Levine also criticized the voluntary "ethical restraint" commitments Anthropic and five other companies signed with the White House, saying they were not enough to address the risks involved. He asked Anthropic to spell out how audit and control measures would be implemented, what areas those measures would cover, and the expected timetable, and requested a written response from the company.
The letter also cited media reports on Anthropic’s draft S-1 filing with the U.S. Securities and Exchange Commission. That draft reportedly warned that advanced AI could pose catastrophic, even existential, risks to humanity, with risk disclosure spanning 80 pages. Levine further raised concern that the company could adopt a governance structure that concentrates control in the hands of founders and trustees. In his view, the gap between control and accountability deserves close scrutiny from investors.
Anthropic and the comptroller’s office are now in contact
Bloomberg Law reported that Anthropic has opened a line of communication with Levine’s office. In a Friday interview with Bloomberg Television, Levine said the company appeared to be "under pressure from multiple directions," but he did not describe the substance of those conversations.
Current direct exposure is about $7 million
According to Agence France-Presse, as carried by Yahoo Finance, the comptroller’s office said New York City funds currently have only about $7 million of exposure to Anthropic.
That figure could rise if Anthropic goes public and is added to indexes tracked by New York City investment programs, including the Russell 3000. In that case, the funds would become long-term shareholders through index funds on a passive basis.
IPO timing and valuation reports remain unofficial
The New York Times reported that Anthropic gave investors a presentation on Oct. 14, saying third-quarter revenue rose about 45% from the previous quarter to $16.8 billion, and that the company could list as early as November.
Underwriting banks said the deal could raise more than $100 billion, implying a valuation of about $2 trillion. At that size, Anthropic could surpass SpaceX, which went public in June this year, and become the largest IPO globally.
IFR, citing people familiar with the matter, said Anthropic planned to publicly file with the SEC on Oct. 26, begin its roadshow on Nov. 9, and price the deal the following week. Reuters separately reported that some investors had lowered their valuation expectations because third-quarter revenue growth was well below the more than doubling pace seen in the second quarter.
Anthropic has not formally announced the timetable or valuation described in those reports, and those details remain based on anonymous sources.
Risk language in the filing has now become a governance demand
New York City’s direct position in Anthropic remains small at roughly $7 million. Still, Levine’s letter shifts the focus from portfolio size to governance: it turns the AI risk disclosures already described in the company’s own listing materials into a specific demand for pre-IPO answers on oversight, controls, and timing, along with a written response.
With Reuters reporting that some investors have already cut valuation expectations, and IFR describing a roadshow that could begin on Nov. 9, Anthropic’s response to the comptroller’s request is now a central point to watch.