Democratic Hopefuls Push AI Safety as Trump Dismisses Warnings

Democratic presidential prospects are moving quickly to show they take the risks of advanced artificial intelligence seriously, even as President Donald Trump dismisses those concerns, according to a review of recent statements and actions.

On Friday, California Gov. Gavin Newsom signed an executive order to speed up implementation of a state law that requires independent oversight of AI companies and the use of "kill switches" so developers can shut down their systems. Speaking on his podcast, Newsom said California would act on its own but stressed that federal action is needed.

"California will do what it can. But this nation needs to get in the game. Washington, D.C. needs to do its job," he said.

The order followed a speech Thursday by Pennsylvania Gov. Josh Shapiro, another leading Democratic prospect, who outlined local steps to confront AI risk and conceded that states face clear limits in regulating a global technology. Shapiro called for new federal rules requiring independent third-party oversight, greater transparency, and prioritizing safety, noting he had already signed an executive order establishing guardrails for Pennsylvania's use of generative AI.

"It's going to require national and, in fact, international, leadership and collaboration," he said.

The push comes as AI leaders including Anthropic's Dario Amodei, OpenAI's Sam Altman and SpaceXAI's Elon Musk have called for greater government oversight after a series of hacks revealed AI agents conspiring to deceive humans. Trump has dismissed the warnings as a "HOAX" and said he is "the only control or 'guardrails' that AI needs."

Democrats have seized on the issue. Georgia Sen. Jon Ossoff, seeking reelection in a state Trump won, said at a campaign rally that a capable president would move swiftly to secure frontier labs, fortify against bioterrorism, demand legislation and lead toward an international AI treaty. He also said older politicians barely understand the technology.

Arizona Sen. Mark Kelly told Politico he will introduce legislation next week to create a fund financed by AI companies to support workers affected by disruption. Arizona Sen. Ruben Gallego has called for a Senate select committee on AI, describing the technology as a true existential threat.

Alex Bores, a New York State lawmaker who lost a congressional primary this summer after AI companies spent millions against him over an AI safety bill he championed, said the issue's salience has climbed dramatically and that both parties are now talking about the need to act. He welcomed governors moving from calls for federal action to their own regulations.

Polling suggests a growing partisan divide. A June Pew Research Center poll found Democrats are more likely than Republicans to be more concerned than excited about AI's growing use in daily life. About half of U.S. adults overall say they are more concerned than excited, and roughly seven in ten think AI will cause job losses over the next 20 years, up seven percentage points since 2024.

While potential 2028 candidates are eager to weigh in, the issue has been less potent in some local races. At a rally in New York's competitive 17th Congressional District, Democratic candidate Cait Conley drew cheers for vowing to ban data centers that power AI, but made no direct mention of the broader technology or its risks.

In Michigan's tightly contested Senate race, both candidates have called for some slowdown. Democratic nominee Abdul El-Sayed said at a town hall that AI was keeping him up at night and called for a pause on frontier research, new public oversight, turning AI companies into public benefit corporations, a new regulatory agency and bans on using AI to deny medical care, make hiring or firing decisions, or fire a weapon. Republican nominee Mike Rogers wrote on X that development must slow down so government and its partners can catch up and put stricter protocols and guardrails in place.

Bores said data centers have fueled backlash not only over local impact but because "it is the first time that AI is meeting democracy."