Connecting Alaska to the World And the World to Alaska
Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations

Why Steve Bannon sided with Sanders, not Trump, on AI

Former White House chief strategist Steve Bannon speaks during the Pro-Human Assembly at the Marriott Marquis Washington on Sept. 15 in Washington, D.C.
Finn Gomez
/
Getty Images
Former White House chief strategist Steve Bannon speaks during the Pro-Human Assembly at the Marriott Marquis Washington on Sept. 15 in Washington, D.C.

Updated September 16, 2026 at 7:46 AM AKDT

Steve Bannon — the podcaster, entrepreneur, and provocateur who was an early backer of President Trump — has parted with the president on artificial intelligence.

Bannon shared a stage this week with Sen. Bernie Sanders, I-Vt., and echoed some of Sanders' concerns about AI, even as the two differed on what to do about it. Bannon has railed for years against tech leaders who have become Trump's allies, and demanded a slowdown in AI development to give the government time to put regulations in place. Trump said this week that the only necessary "guardrail" is a "smart president," which was how he described himself.

"This is one we're going to have to disagree on," Bannon told NPR's Steve Inskeep.

In Washington, a once-abstract debate over AI oversight is turning into a much noisier fight as people building the technology sound urgent warnings. Jacob Coxon, a former Anthropic researcher, announced his resignation in early September and raised concerns about increasingly capable AI systems. Anthropic CEO Dario Amodei then called for slowing the development of more powerful models, a position backed by OpenAI CEO Sam Altman and xAI's Elon Musk. Trump has rejected calls for a slowdown, warning that such a move could give China an advantage.

Bannon arrives at a different answer.

He wants regulation, but not a halt to American AI development. And he is deeply skeptical of the motives of technology companies now asking Washington for help, arguing that new rules could give the largest labs a competitive "moat" while allowing them to shift financial risk onto the public.

"This technology has some real upsides to it, but you have to be realistic," Bannon said. He called for immediate executive action while Congress works on longer-term legislation, then saying the government needs time to build a more detailed regulatory system.

His approach to China is more sweeping.

"I want to shut it down," Bannon said.

Bannon said cutting China off from U.S. technology would give Washington time to build a regulatory system, potentially something like the Atomic Energy Commission.

"If you shut down the Chinese Communist Party and AI, which we can do because we give them everything, we can then take our time and set up a regulatory apparatus," he added.

He also called for expelling Chinese nationals who are employed by American AI companies. This would not be a small change. A Carnegie Endowment for International Peace study found that 87 of 100 top Chinese AI researchers who were working at U.S. companies or universities in 2019 were still at U.S. institutions years later. Separate research found that U.S.-China AI collaboration has occurred more often than collaboration between any other two countries since 2018.

How Washington would untangle those ties is now part of the broader U.S.-China relationship. Trump is planning to host Chinese President Xi Jinping on Sept. 24, pending confirmation from Beijing.

Editor's note: Some of the material in this story was not included on the radio version that aired on Morning Edition. Listen to the conversation by clicking the blue button above. 

The radio interview was edited by Reena Advani. The digital version was edited by Suzanne Nuyen.

Copyright 2026 NPR

Steve Inskeep is a host of NPR's Morning Edition and Up First.
Majd Al-Waheidi
Majd Al-Waheidi is a digital editor for Morning Edition and Up First.