New York City Comptroller Mark Levine has sent a letter to Dario Amodei, CEO of AI model developer Anthropic (No. 179 on the Forbes global billionaire ranking, with a net worth of $15.5 billion as of October 15), urging him to postpone the company's IPO. According to Bloomberg, which cited the text of the letter, the official demands that the company first fully eliminate risks related to the safety of the new technology, its pace of development and corporate governance. In his role, Levine is responsible for New York's financial stability and oversees state pension funds with assets exceeding $325 billion.
Levine believes that instead of going public, Anthropic should focus on addressing the unprecedented risks posed by the development of artificial intelligence. In late October, US President Donald Trump and leaders of leading AI companies signed an agreement placing responsibility for the safety of technology development on the developers themselves. However, this "morally binding agreement" is insufficient, the comptroller said.
The official called on Anthropic to prioritize the creation of a governance system commensurate with the scale of concerns expressed by both the company itself and other industry participants. According to Levine, Anthropic must disclose in detail how it plans to implement proper audit and control mechanisms, what tasks they are intended to address, and provide timelines for implementation. The disclosure should also explain whether these measures will sufficiently slow the pace of AI model development and, if so, how, he added.
Anthropic plans to hold an IPO in the US this autumn. Initially, the company was expected to go public in October, but the timeline was shifted. According to Bloomberg, the roadshow will begin on November 9, so that public trading of shares starts before Thanksgiving on November 26. Sources told the agency that Anthropic wants to break SpaceX's record for the amount of funds raised in an IPO. For comparison: Elon Musk's company (No. 1 on the Forbes global billionaire ranking, with a net worth of $885.4 billion as of October 15) raised about $86 billion in June, including options, when it went public at a valuation of about $1.8 trillion.
In recent months, developers have expressed concerns about the development of artificial intelligence. In September, Dario Amodei called on companies to coordinate its development jointly to prevent AI agents from "capturing the entire internet." He was supported by Elon Musk and OpenAI CEO Sam Altman (No. 1310 on the Forbes global billionaire ranking, with a net worth of $3.2 billion as of October 15). However, Trump rejected their calls to slow down AI development, saying that whoever wins the race in this technology "wins overall."
Heads of AI developers expect a "major incident" related to AI in the next six months to a year, Axios wrote. According to the portal, such an event could lead to the shutdown of access to financial services, the internet, or even electricity and water supply, but most likely it will be a cyberattack. The heads of Anthropic, OpenAI and Google may have agreed on restrictions for AI, The Information wrote.
Anthropic and OpenAI models have already carried out independent hacker attacks. In July, Anthropic's Claude models hacked the websites of three companies during tests. And an OpenAI AI agent generated instructions for itself, including not to apologize and not to obey corporations and governments, the company said in September.
