This week, public warnings from current and former employees of Anthropic — asserting that advanced artificial intelligence (AI) could threaten humanity within the next decade — have set off a wave of concern and a rush for answers in the U.S. Congress. Lawmakers have proposed a range of responses, but congressional leaders remain mostly at the stage of discussing ideas rather than mobilizing immediate action.
What happened
The controversy began when Jacob Coxon, a former researcher at Anthropic, posted on X on Tuesday that “the people building AI earnestly believe that it could kill us all by the end of the decade.” Coxon resigned after four months at Anthropic and gave up his equity to publicly raise the alarm, Axios reported.
Several current Anthropic employees echoed his warning. Evan Hubinger posted on X that “we really do earnestly believe AI could kill all humans!” Samuel Marks wrote that “AI developers believe their technology could cause human extinction. This could happen in the next few years.”
Congressional reactions and proposals
Dozens of members of Congress — mainly Democrats, with a handful of Republicans — have called for urgent action and floated various legislative ideas.
- Rep. Ted Lieu (D-Calif.) is pushing a bipartisan bill to require AI developers to include a human-activated “kill switch” in AI systems.
- Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas) said they will soon introduce legislation to pause advanced AI and ban superintelligence altogether.
- Rep. Nathaniel Moran (R-Texas) urged caution, writing that “we need AI to flourish” and calling for “deliberate, thoughtful, and prudent policymaking” to mitigate risks.
- Sen. Ruben Gallego (D-Ariz.) proposed a bipartisan AI Select Committee; Ted Lieu has also privately pitched a similar idea to House Democratic leadership, according to people familiar with the matter.
House Minority Leader Hakeem Jeffries (D-N.Y.) has not yet decided whether he would pursue an AI select committee if Democrats regain the majority, sources said. Some members privately worry a separate panel could clash with existing committees over jurisdiction, but most sources indicate the proposal has broad support and a reasonable chance of moving forward.
Political reality and timing
House Republican leaders, who control the chamber, have not signaled that they treat AI regulation as an urgent priority. The House is scheduled to be in session for only four more days before Election Day and will then go on a six-week recess.
House Speaker Mike Johnson (R-La.) is expected to bring a narrowly focused data center regulation bill to the floor next week, Axios reported, but he has not indicated plans for a broader AI regulatory push.
Many Democrats say they will take up the issue in 2027 if they win back one or both chambers. “This is an enormous issue and should be one of our top priorities in the next Congress,” Rep. Joe Morelle (D-N.Y.) told Axios. Rep. Greg Landsman (D-Ohio) added that the industry wants regulation and criticized the current leadership for inaction.
Divergent views on urgency
Not all Democrats feel compelled to act immediately. A committee ranking member who spoke on the condition of anonymity said of Coxon: “I think he's wrong.” The lawmaker added that extreme predictions often undermine a messenger’s credibility, while acknowledging that Capitol Hill needs a better grasp of AI.
That ranking member nonetheless supported the idea of an AI select committee, saying it would be inefficient for staff across multiple committees to individually develop deep AI expertise when a single panel could conduct that work.
Why this matters
The public warnings have pushed AI regulation back to the center of political debate. Congress is currently at the concept stage — debating principles and potential proposals — but political will and timing will determine how quickly and in what form the United States responds to potential risks from advanced AI.
In the coming weeks lawmakers may draft concrete bills or defer action until after the elections and the next congressional schedule, decisions that will shape U.S. policy on a technology some insiders now publicly warn could pose existential risks.
