New York City Comptroller Mark Levine is urging Anthropic to reconsider its planned initial public offering next month, warning that going public could distract the AI company from its safety work as concerns about AI-related risks grow.
Levine told CNBC’s Squawk Box on Thursday that an IPO would take time and could put Anthropic under pressure to make decisions that clash with its safety priorities.
"This will undoubtedly distract them from the safety work," Levine said Thursday on CNBC. "Once they are public, they are going to have new pressures on growth and profits that are going to push against the safety work, and we have skin in this game in the New York City pension funds because we track to indexes."
On Wednesday, Levine sent a letter to Anthropic CEO Dario Amodei pressing the company to pause its plans to go public next month amid growing safety concerns and a growing divide between some AI industry leaders and the Trump administration over how the technology should be developed.
A researcher, who previously worked at both Anthropic and its rival OpenAI, has publicly accused both companies of moving too quickly and putting humanity at risk. Since then, both CEOs of the company have supported calls for safeguards and said a slower pace of development is essential.
Meanwhile, the Trump administration has taken a different route amid its race to get ahead of China. Earlier this month, Trump renamed AI to "super intelligence" and created a task force that is tasked with coordinating the federal government's AI-related engagement with consumers, public interest groups, religious organizations, critical infrastructure providers and AI companies.
Over the past month, AI leaders have met with Trump and signed a voluntary agreement that includes partnering with an outside auditor and having internal controls in place. Levine, however, said those measures fell short and are not enough to "meet the moment."
"Considering these concerns, Anthropic should disclose with specificity how it will implement appropriate audits and controls, the scope of what these controls are intended to address, as well as a projected timeline," Levine said. "This should also include an explanation of whether and how these controls will sufficiently address your call to slow the pace at which the capabilities of AI models advance."
Anthropic did not immediately respond to a request for comment.
In his role as comptroller, Levine also serves as an investment adviser to the city's pension funds, which is one of the largest in the U.S., with $325 billion in assets, according to his letter sent to Anthropic.
"So as future … investors, we care about this and we think we should slow this down until they have a better grip on the safety," Levine said.