After an Anthropic researcher resigned and warned that artificial intelligence (AI) could wipe out humanity by 2030, bipartisan lawmakers are urging swift congressional action.
After spending years doing research for OpenAI and transitioning to Anthropic earlier this year, Jacob Coxon said he resigned from the latter on Tuesday, claiming that “neither company is acting responsibly.”
Coxon warned that those building AI sincerely believe the technology could kill all of humanity by the decade’s end, and said that it’s not a “marketing stunt.”
Several other leading AI researchers, including Anthropic’s Evan Hubinger, agreed with Coxon that the technology could pose a doomsday threat if left unchecked.
“We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger wrote on X. “I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”
Lawmakers from across the aisle urged Congress to take immediate action and apply guardrails to the technology before humanity loses control.
“The call is coming from inside the house,” Rep. Lori Trahan (D-Mass.) wrote on X, calling on Congress to unite and pass her bipartisan FRONTIER Act with Rep. Jay Obernolte (R-Calif.), which would institute a national “risk-based framework” guiding the creation and deployment of the most advanced AI models.
“Safety researchers are resigning, powerful AI models are breaking out of their labs, and companies are racing ahead anyway,” Trahan said in her post.
On the other side of the aisle, Rep. Anna Paulina Luna (R-Fla.) on Wednesday urged Congress to convene a special session on technology and the nation’s future.
“Partisan politics aside, there are massive implications of a race towards super intelligence,” Luna wrote. “Our first priority is protecting ALL Americans. To include data, privacy, and physical safety. We MUST be ready. This is happening whether or not we want it to.”
Responding to Coxon’s warnings on social media about technology’s dangers, Rep. Ted Lieu (D-Calif.) called for passing the “bipartisan AI Kill Switch Bill asap.”
Lieu is co-chair of a House AI commission tasked with developing AI policy recommendations by Minority Leader Hakeem Jeffries (D-N.Y.).
OpenAI and Anthropic did not respond to requests for comment by publication time.
Anthropic CEO Dario Amodei penned an op-ed in The New York Times in June 2025 calling on Congress to put limits on AI development.
Lawmakers were not the only ones calling for action, urging the U.S. government to take AI threats seriously and formulate plans to restrain the technology.
A group of bipartisan organizations released a joint letter on Wednesday urging President Donald Trump to publicly release the framework on government AI testing that his administration developed privately earlier this year.
“The government has a fundamental responsibility to share the parameters of the process for reviewing frontier AI models prior to wider deployment,” the organizations wrote in the letter led by the Center for Democracy and Technology and Americans for Responsible Innovation, a nonprofit that advocates for stronger regulations on the AI industry.
“Through this framework, the federal government will obtain initial access and visibility into the most transformative technologies we have ever seen.”
In June, Anthropic called for a worldwide slowdown in AI development, warning that the models are reaching a point at which they could soon improve themselves independently, potentially taking control out of humanity’s hands.
“This is called recursive self-improvement,” Anthropic’s Marina Favaro and Jack Clark wrote in a blog post. “We are not there yet, and recursive self-improvement is not inevitable. But it could come sooner than most institutions are prepared for.”
Weeks later, a partnership of intelligence agencies from the United States, the United Kingdom, Canada, Australia, and New Zealand warned that AI models could breach and overwhelm the cybersecurity defenses of governments and businesses worldwide within months.
“Frontier AI models are anticipated to exceed current industry expectations, fundamentally transforming both offensive and defensive cyber capabilities. The timeline is not years, it is months,” the Five Eyes Intelligence Oversight and Review Council statement said.
We had a problem loading this article. Please enable javascript or use a different browser. If the issue persists, please visit our help center.