Sen. Mark Kelly is skeptical that artificial intelligence will kill everybody in the next 10 years, but he’s worried about what the fast-growing technology will do to the American workforce.
Kelly, an Arizona Democrat and one of his party's 2028 presidential prospects, has a new plan to tax big AI companies and put the cash in a fund to help workers who lose their jobs due to the new technology.
The proposal comes after AI leaders sounded the alarm that the technology is accelerating at a dangerous — and potentially deadly — pace.
“The CEOs of these companies are asking for the federal government to regulate them and consider the safety of these models. There's no reason why we shouldn't act,” Kelly told The Arizona Republic in a Sept. 23 telephone
interview. “We should take steps to, you know, mitigate the potential of this technology doing harm to large numbers of people.”
Kelly planned to file his new bill, the Make AI Work for Americans Act, on Sept. 24. He’s one of several Democrats considering 2028 presidential campaigns who are laying out their plans for how to tackle AI.
Kelly’s legislation would create an “AI Horizon Fund” that would be supported by taxes on AI companies. The legislation would also upgrade state unemployment systems, use grants to create pathways into jobs that “rely on human judgment” and set up new programs for entrepreneurs.
Kelly said it’s not clear how many people will lose their jobs to artificial intelligence in the coming years. So far, losses have fallen short of the most pessimistic estimates, he said.
“That doesn't mean things can't change,” Kelly said. “I don't think anybody knows, and I think if they tell you it's going to be this, you know, they're probably somewhat full of (expletive).”
The proposed AI Horizon Fund would launch with a $30 billion advance appropriation that would later be paid back as the industry pays into the fund. Money for the fund would come from a digital advertising tax, a usage tax based on output from AI models and an excess profit tax on companies with particularly high revenue.
Kelly has spoken with leaders at OpenAI and Anthropic, who he said are on board with regulation. His team has been in touch with just about every big AI company, he added.
“They realize that their technology might be so disruptive that there's going to have to be a plan in place for workers. And in my conversations with them, they get it, and they think they should be part of the solution here,” Kelly said.
Kelly also took a shot at President Donald Trump, who has pushed back on calls to regulate AI.
"The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!" Trump, a Republican, wrote on Truth Social. "We already have tremendous CRIMINAL and REGULATORY power over these companies!"
Although Kelly didn’t use AI to write his new bill, he said he uses OpenAI’s ChatGPT and Anthropic’s Claude to do research and sort through data. He doesn’t have a favorite between the two chatbots.
Lawmakers and others have proposed a flurry of policy proposals for AI in recent weeks in reaction to the chilling warnings about the technology. Kelly said he plans to touch base with former White House Press Secretary Jay Carney, who earned the praise of former Vice President Kamala Harris for his work on an initiative to connect policymakers with AI firms through his role at venture capital firm SV Angel.
However, Kelly threw cold water on the most dire warning about AI, which came when Anthropic researcher Evan Hubinger said on X that “we really do earnestly believe AI could kill all humans!”
That seems a little extreme to Kelly, who pointed to his past as an astronaut. Still, regulation of AI is appropriate even if the fear of catastrophe is far-fetched, he said.
“I used to fly the space shuttle, and we would talk about risk a lot. You know, some risk could be minor; other things could be catastrophic, and it's an area of engineering called probabilistic risk assessment,” Kelly said. “When something is catastrophic, you don't ignore it, even if the probability is low.”
Stephanie Murray covers national politics and the Trump administration for The Arizona Republic and azcentral.com. She co-hosts “The Gaggle,” The Republic’s weekly politics podcast. Reach her at stephanie.murray@usatodayco.com, on X and TikTok @stephanie_murr and on Instagram @stephaniemurray.news.
This article originally appeared on Arizona Republic: Will AI kill us? Mark Kelly doubts it, but worries it could kill jobs | Exclusive
