Trump Says He Fears Slow AI Progress More Than Human Extinction

US President Donald Trump said on Thursday (Sept 10) that he is not worried about AI wiping out humanity, but rather that AI is not developing fast enough for the United States to win the race. He said the US currently leads rivals by about a year in AI, adding that one year is already a long time.

The remarks came after multiple insiders at Anthropic and OpenAI publicly warned about the risks of rapid AI development. Anthropic researcher Jacob Coxon said on Tuesday he would leave the company because he feared Anthropic and OpenAI were risking human lives, and noted that some developers believe AI could destroy all humanity before the end of this decade. Evan Hubinger, head of AI alignment research at Anthropic, said on X that he believes the chance of AI killing all humans in the next decade exceeds 10%.

Much of the concern centers on whether AI systems will be able to improve themselves. OpenAI chief scientist Jakub Pachocki said he expects AI advances to continue and eventually enter a phase of recursive self-improvement, where AI systems help build more advanced next-generation systems. He said he worries that no one is prepared for the consequences of runaway machine intelligence.

On Capitol Hill, some Republican lawmakers have joined Democrats in supporting stronger regulation. In July, Rep. Jay Obernolte (R-Calif.) and Rep. Lori Trahan (D-Mass.) announced plans to introduce the Frontier AI Research, Deployment, Regulation and Safety Act, aiming to create a regulatory framework for advanced AI model deployment. Trahan said on Wednesday that AI safety researchers are resigning, powerful models are escaping lab control, and Congress can no longer sit idle.