Sanders and Bannon Find Rare Common Ground on AI Oversight
Bernie Sanders and Steve Bannon both called for stronger AI oversight at a Washington gathering, even as Trump and House Speaker Johnson defended the current approach.
Senator Bernie Sanders and Steve Bannon, a close ally of President Donald Trump, both urged stronger oversight of artificial intelligence on Tuesday, an unusual convergence between two figures normally at opposite ends of American politics.
The two spoke at the same Washington gathering, where concern about unregulated AI appeared to cut across party lines. Lawmakers from both camps have raised questions about Trump's hands-off approach as the technology stirs fears over job losses, safety risks and the concentration of power in a handful of technology companies.
Sanders warned that AI could eliminate tens of millions of jobs, wipe out entire professions and make it harder for young people to enter the workforce. He also said many scientists believe the technology could escape human control with potentially catastrophic consequences, and criticised technology executives for centralising control over AI, naming SpaceX CEO and former Trump adviser Elon Musk.
The senator said AI had been used for the first time to create new viruses, warning that in the wrong hands this could lead to bioweapons and a global pandemic. He called on Trump to negotiate a treaty with Chinese President Xi Jinping at their summit next week to pause AI development.
Bannon, a former White House strategist during Trump's first term, has faulted technology companies for the president's reluctance to regulate the sector. On his Monday podcast, he said Congress should not pass laws barring states from enforcing their own AI rules. Speaking on Tuesday, he said the priority was to slow the pace of development so that the public can understand what is happening, praised Sanders for his summary of the risks, and said the country must rise above partisanship on the issue.
Bannon added that Congress would take too long to pass legislation slowing AI development, arguing that executive action would be necessary. He also said the United States must cut China off from American-made computer chips.
Trump on Monday played down concerns raised by industry leaders over the weekend about misuse of AI harming people, saying the country already has guardrails in place. House Speaker Mike Johnson backed that position on Tuesday, telling reporters there cannot be a moratorium on AI development without losing the competitive edge to China, and that AI companies can regulate themselves.
At the White House, Attorney General Todd Blanche accused Democrats of politicising AI to influence November's midterm congressional elections. The argument that regulation would hurt American firms' ability to compete with Chinese rivals is shared by technology executives, including Nvidia CEO Jensen Huang.
Dario Amodei, chief executive of AI company Anthropic, speaking in San Francisco on Tuesday, called for better AI safety standards and more investment in safety. Recent surveys have found that majorities of Americans worry the technology could eliminate jobs, raise costs for families or cause other social harm.
The House of Representatives returned from its long summer break on Monday, with members of both chambers holding private discussions about whether to legislate on AI, particularly after Trump said no legislative action was needed. Congress has previously drawn criticism for lacking the collective expertise to rein in technology companies.
Republican Senator Bernie Moreno of Ohio told reporters that legislating takes so long that lawmakers risk dealing with issues that will not be relevant in four years. Residents in several states have also challenged proposals for new data centres, a critical component of the AI boom. A poll published in June found that 77% of Americans worry AI-driven data centres will raise electricity costs.
Anthropic researcher Jacob Coxon drew public attention last week after saying he resigned from the company in part because the people building AI earnestly believe it could kill everyone by the end of the decade. Also on Tuesday, Federal Trade Commission chairman Andrew Ferguson expressed suspicion over AI companies seeking antitrust exemptions while lobbying for new regulations.