President Donald Trump said Thursday he has no concerns about artificial intelligence causing human extinction, dismissing warnings from researchers at two leading AI labs. Speaking to reporters, Trump said his main focus is winning the AI race against China, adding that he believes the United States is currently ahead by about a year.
His comments came after a wave of public safety warnings. On Tuesday, Anthropic researcher Jacob Coxon announced his resignation, saying Anthropic and OpenAI are gambling with our lives and claiming that people building AI believe it could kill us all by the end of the decade. OpenAI alignment researcher Jasmine Wang wrote Wednesday that it is hard to overstate how dangerous speeding toward recursive self-improvement, or RSI, is. OpenAI chief scientist Jakub Pachocki said Saturday he has a strong expectation that AI progress could continue into RSI territory and is concerned no one is prepared for the consequences.
Lawmakers are moving to increase oversight. In July, Reps. Jay Obernolte and Lori Trahan announced the FRONTIER Act, which would set rules for deploying advanced AI models. Earlier this month, Sen. Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act, which would pause advanced AI development until federal safety rules are in place. Trahan said on X that safety researchers are resigning, powerful models are breaking out of labs, and companies are racing ahead anyway. Trump has not signaled support for either bill.
OpenAI, meanwhile, has asked members of Congress whether rival AI companies could legally coordinate an AI slowdown without violating antitrust laws, according to WIRED. The outreach follows Pachocki's call for voluntary slowdowns. Experts said competitive and financial pressures discourage independent action. Miranda Bogen of the Center for Democracy and Technology said companies are incentivized to release products before risks are fully understood. Duncan Sabien of the Machine Intelligence Research Institute said each advance yields millions or billions in funding, and an arms-race dynamic means stepping back just gives a rival the lead.
Trump delayed an AI executive order in May over concerns it could weaken America's lead over China, then signed it in June, creating a voluntary review process for advanced models before release. In August, OpenAI paused internal Astra work over cybersecurity safeguards, while both OpenAI and Anthropic relaxed safety commitments in February. Coxon said his resignation was not a marketing stunt, and many executives express fear privately. Sabien argued researchers may accomplish little by leaving unless enough of them stop together to break the coordination barrier.
