NYC Comptroller urges Anthropic to postpone US IPO over safety concerns

Mark Levine, the New York City Comptroller, sent a letter to Anthropic CEO Dario Amodei urging the artificial-intelligence lab to hold off on its US initial public offering until it can adequately address safety, governance and the speed of its development. Levine warned that AI capabilities have progressed faster than regulatory frameworks, creating heightened concerns over safety, cybersecurity threats and potential misuse.

He emphasized that the company should prioritize tackling these risks instead of moving ahead with the listing. Anthropic’s internal research has already pointed to dangerous abilities in its models, including sabotaging code, enabling fraud and manipulating information. The IPO is slated for November and is reported to target a valuation exceeding $2 trillion, while a rival, OpenAI, plans to list later in 2027. The comptroller’s office and Anthropic did not immediately reply to inquiries.