Fears that AI poses an increasing threat to safety and prosperity have made progressive firebrand Senator Bernie Sanders and conservative provocateur Steve Bannon unlikely allies with their demands for more safeguards. Both men appeared Tuesday at an event in Washington where each warned that the technology carries potentially catastrophic risks and that AI companies have amassed vast power, jeopardizing human interests.
Sanders and Bannon each took aim at AI billionaires for moving too quickly on the technology. But Sanders favors congressional action — a time-consuming endeavor that would require bipartisan support — while Bannon would prefer President Donald Trump take executive action.
"The legislation's going to be enormously complicated and take a while, just because the technology's complicated and the opportunities are complicated and the bottomless pit abyss of the downside is complicated," Bannon said.
The daylong gathering, whose attendees also included investor Frank McCourt and Massachusetts Institute of Technology economist Daron Acemoglu, took place against a backdrop of intensifying AI anxiety, touched off by warnings from Silicon Valley researchers that cutting-edge systems carry an increasing risk of slipping beyond human control. Those concerns have fueled calls for lawmakers and the Trump administration to address AI risks, though there's little sign of any imminent action — which Sanders lamented in his remarks.
"If AI surpasses human intelligence, as many scientists believe could happen, this technology will escape human control with potentially catastrophic consequences," said Sanders, a Vermont independent who aligns with Democrats. "Congress, under both Democratic and Republican control, have been asleep at the wheel and we now have a president whose ignorance regarding this issue is truly embarrassing."
Bannon, the architect of Trump's 2016 election victory, took the stage right after Sanders and praised the senator for doing "a great job of summarizing the risk."
"Obviously he and I have a quite different opinion of the president of the United States but we're trying to raise up above partisanship and we have to," he added.
Debate over addressing AI's risks was galvanized over the weekend by a 3,800-word essay from Anthropic PBC Chief Executive Officer Dario Amodei, who promised to rein in development of the most advanced systems so researchers can better understand potential threats. OpenAI CEO Sam Altman and SpaceXAI CEO Elon Musk — whose companies are bitter rivals to Anthropic — quickly endorsed Amodei's missive, adding to a groundswell across Silicon Valley supporting steps to address AI's increasingly unpredictable perils.
Amodei cited two main reasons for slowing AI development: the technology's increasing ability to improve itself and a July incident where a swarm of AI agents from OpenAI went undetected as they collaborated to breach the Hugging Face Inc. website. In his essay, the Anthropic chief urged policymakers to enact regulation to promote transparency among AI developers and encourage outside audits of new systems.
For now, the Trump administration is resisting calls for further AI rules. On Monday, Trump belittled the push by Anthropic and OpenAI to slow work on cutting-edge systems, dismissing existential fears as "a hoax." In a series of bombastic social media posts, Trump squashed the idea of new federal guardrails on the technology, citing AI's importance to the domestic economy and in preserving the US lead over China in technology. With the president rejecting tighter oversight, US lawmakers are floating a range of measures aimed at addressing AI safety concerns. Sanders said Tuesday that he plans to introduce legislation next week that would ban the so-called superintelligent AI — with capabilities potentially surpassing human control — that stirred Amodei's warning.
Risking Trump's wrath, even Republican House Speaker Mike Johnson has indicated his openness to regulation, saying that lawmakers "potentially" will need to impose guardrails, though he said it would take "some time" to fashion legislation with enough support to pass. On Tuesday, he cautioned against overreach in enacting. "We cannot have a moratorium on the development of AI," he said. "We will lose the race."
Odds of any proposal passing before the end of the year remain slim. Congress is unlikely to act on any AI legislation anytime soon, with the House slated to leave Washington on Thursday and not return until after the November elections. The brief lame duck session after the election is expected to focus on a plan to boost the Pentagon budget.
At Tuesday's conference, speakers throughout the day highlighted a wide range of AI risks, including those close to home. During a panel on children's safety online, two parents aired their stories about the harmful and misleading conversations their children had with chatbots before they killed themselves.
In another session, actress and activist Ashley Judd discussed ways AI can reflect bias against women. Experts with the Future of Life Institute, the group that organized the conference, spoke about the potential risks of automated weapons and superintelligence without human oversight.
Both Sanders and Bannon assailed the concentration of power among Silicon Valley billionaires. "Virtually all the decisions regarding the development of AI have been made by a handful of the richest people in the world," Sanders said.
For his part, Bannon accused the AI CEOs of trying to turn a crisis into their advantage. "What they want is to socialize all the risk with the American people, including potential bailouts when they come, and they want to keep all the upside," he said.
While Sanders hailed the pledges from Anthropic and OpenAI to slow their work on advanced systems, he said that is "a start, but it's not enough."
"When you are racing towards a cliff, you don't just ease up on the gas. You hit the brakes," Sanders said.