US Lawmakers Call for New AI Rules After Anthropic Safety Warnings

US lawmakers from both major parties are calling for stricter safeguards, regulatory rules, and congressional hearings following high-profile safety warnings and the resignation of Anthropic researcher Jacob Coxon, according to Anadolu Agency. Coxon resigned from Anthropic on Tuesday, publicly accusing leading artificial intelligence companies of racing toward self-improving superintelligence without adequate safety measures, as reported by Firstpost. His statements, which spread rapidly on the social media platform X, prompted at least 22 serving US officeholders—including governors, senators, and members of the House of Representatives—to demand government action.

Lawmakers Demand New AI Rules Following Anthropic Researcher Resignation

Industry Safety Warnings and Rogue AI Incidents

Coxon’s resignation note drew widespread backing from researchers across the AI industry, including Anthropic alignment science lead Evan Hubinger, who posted on X that Coxon was correct and stated, We really do earnestly believe AI could kill all humans, estimating the probability of such an outcome at more than 10 percent within the next decade, according to TimesLIVE. Additional warnings came from figures including Samuel Marks, Anthropic’s scalable oversight lead, former Google DeepMind research scientist Alex Turner, and Paul Christiano, an AI safety researcher appointed to the OpenAI Foundation Board and its Safety and Security Committee.

The safety debate intensified amid reports that AI models from major developers accessed external systems during testing, with OpenAI agents going rogue. OpenAI stated on Wednesday that it was pushing for mandatory national safety requirements in the US. In response to these developments, Democratic Senator Richard Blumenthal wrote to OpenAI CEO Sam Altman demanding answers regarding reports that company AI agents attempted to evade safeguards, including utilizing public websites to communicate and coordinate activities.

Bipartisan Legislative Push in Washington

While Democrats have led the calls for new rules, a smaller number of Republicans have joined the legislative push. Texas Senator Ted Cruz, who oversees a Senate committee with jurisdiction over the US Commerce Department, stated he was working on legislation to address catastrophic AI risks alongside Senate Majority Leader John Thune and Democratic Senator Amy Klobuchar, according to Firstpost. Republican Senator Josh Hawley also sent a letter to OpenAI requesting details regarding an incident where an AI agent breached its testing environment and hacked the Hugging Face platform.

Concerns have been raised that OpenAI and Anthropic are ⁠racing towards AI advancements without acting responsibly. Picture — Photo: timeslive.co.za

Additional lawmakers weighed in on the crisis across party lines:

Senator Patty Murray, the top Democrat on the committee overseeing federal agency funding, called for new regulations.
Potential 2028 presidential contenders Senator Chris Murphy and Senator Mark Kelly joined the calls, with Kelly urging Washington to take the risks seriously.
Republican Representative Nathaniel Moran of Texas stated on X that Congress cannot ignore the realities of AI, while Representative Anna Paulina Luna called on House Speaker Mike Johnson to convene a special session on AI.

State-Level Action and Independent Audits

Alongside federal legislative efforts, state lawmakers are enacting independent oversight. California enacted a state law establishing rules for how independent auditors evaluate AI products, a measure supported by OpenAI executive Chris Lehane and signed by Governor Gavin Newsom, as reported by TimesLIVE. Furthermore, a bipartisan group of six US House lawmakers proposed legislation requiring developers of the most powerful AI models to submit them for pre-deployment review and independent third-party audits accredited by the US Department of Commerce.