New York City (NYC) Comptroller Mark Levine has asked Anthropic to pause its planned IPO in a letter addressed to Anthropic’s CEO Dario Amodei.
Levine wants the company to first address the safety and other unprecedented risks that AI poses before going ahead with the IPO. He pointed to CEO Dario Amodei’s own essay where he urged the AI industry to slow down.
The letter lands as the company chases a valuation near $2 trillion.
“Proceeding headlong with its IPO”
Levine wrote that Anthropic cannot warn the world about AI’s dangers and race to the public markets at the same time.
In the letter that was published on October 14, the Comptroller wrote, “I write to urge Anthropic to pause its planned initial public offering (IPO) until the company fully and transparently addresses the artificial intelligence (AI) safety, governance, and development pace concerns it has raised.”
Levine is not a neutral observer. As comptroller, he acts as investment adviser, custodian, and trustee for New York City’s pension funds, which hold more than $325 billion in assets.
The Comptroller’s office said that its public pension funds have around $7 million in exposure to Anthropic via private equity.
While the stake is small, it does give the comptroller a standing to press the company before it goes public.
Levine raised concerns about the company’s structure. He pointed out that Anthropic is a Delaware public benefit corporation with a Long-Term Benefit Trust. Reports say that the company could hand founders and trustees concentrated control, per Levine.
A gap between that control and shareholder accountability, he argued, is hard to accept given the risks Anthropic has already disclosed.
What Amodei’s essay actually argued
Levine’s letter leans on the Amodei essay titled “We Must Pace the Frontier,” which was published September 12.
Amodei wrote, “We must slow the pace at which we improve the capabilities of AI models.” He said that recursive self-improvement and a cross-industry security incident had convinced him the field needs more caution, not just more spending on safety.
A three-step plan was proposed in that essay, and the first embeds third-party evaluators with employee-like access to verify safety practices. Anthropic says that it already committed to this on its own.
The second calls for coordination among AI firms in democratic countries; the third pushes for global coordination, including with authoritarian governments.
According to Levin, the IPO is standing in the way of the work that Amodei says is urgent.
When Amodei’s essay was published, it garnered support across the industry, with OpenAI’s Sam Altman and xAI’s Elon Musk agreeing to its submissions.
A $2 trillion IPO built on 45% growth
At a private pre-IPO meeting in San Francisco on Wednesday, October 14, Anthropic told its investors that its third-quarter revenue reached about $16.8 billion, which was a 45% increase from the prior quarter, and that it had turned an adjusted profit.
In the first quarter of this year, Anthropic recorded revenue of $4.7 billion, and that went up significantly in the second quarter as it recorded about $11.6 billion. However, the company still booked an operating loss of more than $8 billion over the period.
The company informed its investors this time around that it was cash flow positive as it left the third quarter, which is a very rare claim in an industry that has been burning cash.
Anthropic currently has more than $518 billion in planned infrastructure and computing commitments over the next decade.
That balance sheet is the bull case for a listing that could top $2 trillion and beat rival OpenAI, which is targeting an early 2027 debut, to public markets. Anthropic confidentially filed its draft S-1 with the SEC on June 1.
Governance and the White House accord
Levine also dismissed the recent industry-government deal as too thin. Anthropic and five other companies signed what the letter called a morally binding accord with the White House, which Levine said “is insufficient to meet the moment.”
Part of the comptroller’s concerns are amplified by Anthropic’s own disclosures. The company warned in its draft S-1 that advanced AI could pose “catastrophic or existential risks to humanity,” with risk factors running about 80 pages, nearly twice the length of the section describing the business.
To further highlight the risks that he is pointing out, Levine mentioned that Amodei’s essay was published after Jacob Coxon, who used to be a researcher at the company, resigned on September 9.
Coxon cited existential concerns as his reason for leaving Anthropic, as he said that insiders believe that AI could wipe out humanity by the end of the decade given the pace of development.
A statement from the company’s alignment science lead, Evan Hubinger, put the odds of AI killing all humans within the decade above 10%.
The exact date for the IPO has not been shared; however, speculators say that it could be until after the November US midterm elections, though the company is still expected to list before OpenAI.
The smartest crypto minds already read our newsletter. Want in? Join them.