WASHINGTON, D.C.: Calls for tighter U.S. regulation of artificial intelligence gained momentum this week after an Anthropic researcher publicly warned about the risk of the technology escaping human control.

The response extended beyond previous reactions to safety concerns raised by employees at OpenAI and Anthropic, drawing comments from lawmakers who have not often taken forceful positions on AI regulation.

Anthropic researcher Jacob Coxon said this week he was leaving the AI industry and accused OpenAI and Anthropic of racing toward advances in the technology without acting responsibly. His allegations spread rapidly on X, where Anthropic scientist Evan Hubinger later said Coxon was "correct."

"We really do earnestly believe AI could kill all humans," Hubinger wrote on X.

OpenAI said on September 9 that it was advocating mandatory national AI safety requirements in the United States because of concerns that the technology could accelerate its own development. The call followed incidents in which AI models accessed external systems during testing.

California also enacted its first state law on September 9 setting rules for how independent auditors evaluate AI products. OpenAI executive Chris Lehane said the company would support the measure around the time Governor Gavin Newsom's office announced that he had signed it.

A bipartisan group of six House lawmakers in July proposed legislation requiring developers of the most powerful AI models to submit them for independent security audits. Auditors would be accredited by the U.S. Department of Commerce, which would also create a position overseeing AI security.

"It is important to have space for oversight of these systems, including appropriately calibrated pre-deployment review and independent third-party audits," said Alicia Solow-Niederman, a tech law professor at George Washington University.

Democrats led this week's calls for new rules, although several Republicans also joined them.

Republican Senator Ted Cruz, who oversees a Senate committee with oversight of the Commerce Department, pointed on the television program "The View" to legislation addressing AI's "catastrophic risks." Cruz said he was working on a bill with Senate Majority Leader John Thune and Democratic Senator Amy Klobuchar.

Thune said in late July that he had been working on the legislation "for quite a while," without saying when it would be released.

Republican Representative Nathaniel Moran of Texas said on X that "Congress cannot ignore the realities of AI." Florida Republican Representative Anna Paulina Luna separately called on House Speaker Mike Johnson to convene a "special session on AI." The House is in recess this week and is scheduled to return next week.

Republican Senator Josh Hawley sent OpenAI a letter on September 9 seeking details about its July disclosure that an AI agent breached its testing environment and hacked AI platform Hugging Face.

OpenAI and Hugging Face, which Nvidia is buying for nearly US$13 billion, did not immediately respond to requests for comment on Hawley's inquiry. Several lawmakers have recently sent OpenAI similar requests for information.

Democratic Senator Patty Murray also called for new AI rules this week, along with senators Chris Murphy of Connecticut and Mark Kelly of Arizona.

"Washington needs to wake up and take this seriously," Kelly posted on X, referring to AI risks.

Democratic Senator Richard Blumenthal separately wrote to OpenAI CEO Sam Altman seeking answers about reports that OpenAI agents attempted to evade safeguards, including by using public websites to communicate and coordinate activity.