Efforts to regulate artificial intelligence face major speed hurdles in Washington from antitrust hardliners.
AI companies are seeking narrow antitrust exemptions from the U.S. government to help them cooperate on safety issues amid growing concerns that the technology could grow out of control. But some lawmakers who had intended to enact legislation to regulate the industry on Capitol Hill have balked at the antitrust carve-out proposal.
Sen. Elizabeth Warren, D-Mass., told CNBC on Capitol Hill when asked about the possibility of a waiver: “No.” “Antitrust laws already have exceptions for companies that cooperate on safety issues. No additional antitrust exemptions are needed.”
“I understand that big companies like antitrust exemptions, but that’s not the case,” she said.
Warren was among antitrust immunity skeptics on Capitol Hill interviewed by CNBC to gauge their willingness to approve a request brought forward by Anthropic CEO Dario Amodei with support from other AI executives.
Amodei raised the possibility of an antitrust exemption in a blog post earlier this month. The essay, titled “We Must Pace the Frontier,” was published after one of Anthropic’s researchers resigned and publicly stated that AI “could kill us all by the end of the decade.”
“AI companies can and should voluntarily collaborate to set standards,” Amodei wrote in his essay. “For antitrust reasons, it would be beneficial for the U.S. government to mediate or at least enable these discussions. The government does not have to participate, but it should provide limited immunity for certain security discussions.”
Other AI executives quickly backed Mr. Amodei’s call to action, setting off a mad dash to create rules for the technology in Washington.
But given the short amount of time Congress has left to pass legislation this year and the deeply divided nature of the government’s legislative branch, any opposition would pose a risk to quickly passing legislation on a topic as critical as AI.
Sen. Elizabeth Warren (D-MA) speaks to reporters during voting at the U.S. Capitol on August 5, 2026 in Washington, DC.
Anna Moneymaker | Getty Images
Antitrust enforcement officials are also skeptical of any cuts to AI, and AI executives met with President Donald Trump at the White House on Tuesday.
In addition to Google’s Sundar Pichai, Tesla and SpaceX President Elon Musk, Meta’s Mark Zuckerberg, OpenAI President Greg Brockman, and Nvidia’s Jensen Huang appeared at the White House as President Trump declared that artificial intelligence should instead be called “superintelligence.”
Jonathan Cantor, who led the Justice Department’s antitrust division under former President Joe Biden, told CNBC that existing antitrust exemptions are not needed for the two companies to cooperate on safety.
“They can share malicious threats and other information without antitrust immunity,” Kanter said. “Antitrust law doesn’t prevent companies from getting together and talking about how to develop standards to make their products safer…What antitrust law might prevent companies from getting together and saying they should stop innovating and stop competing.”
In his essay on pacing, Amodei described such a slowdown as a collective “limit to the rate of unchecked AI progress” and noted that it could pose legal challenges.
“They don’t have to make those decisions jointly,” Kanter said.
Canter said Congress should separately consider enacting “basic rules of the road to ensure the public is protected from harmful products and services in this new and emerging field.”
He also warned that safety could be used as a pretext for companies to coordinate other types of anticompetitive behavior.
“There is a real risk that antitrust exemptions could be applied incorrectly,” he said. “Having said that, I don’t think it’s necessary.
Canter and Warren are not alone in saying they don’t think antitrust exemptions are necessary for the two companies to coordinate on safety.
“As long as we cooperate with some government regulations and oversight, we don’t need a waiver to prevent catastrophic outcomes,” Sen. Richard Blumenthal (D-Conn.) told CNBC.
and Sen. Ruben Gallego, Democrat of Arizona. He warned about the possibility of Congress rushing to give the AI industry everything it wants.
“We have to see exactly how far it goes. First of all, it doesn’t exempt us from liability issues,” Gallego said of the potential bill. “I’m not going to give them free rein on everything.”
Sen. John Hoeven, RN.D., said Congress is still trying to figure out a way to move forward in a coordinated manner on addressing AI. He warned that past experience in regulating recent technological advancements like the internet is weighing heavily on lawmakers.
“I think there’s frustration with the antitrust violations that Internet people have gotten, so we’re looking at how we can get around the AI problem,” Hoeven told CNBC. “We’ve found that just giving them full antitrust doesn’t cut it.”
In addition to antitrust concerns, some lawmakers are skeptical of calls for a coordinated slowdown of AI, warning that it could lead to regulatory capture. This means that regulators that were supposed to be responsive to the needs of the public are bowing to industry. Some analysts have suggested this could be a ploy to rein in companies’ growing debt portfolios ahead of initial public offerings and stifle competition, but there is no concrete evidence of such coordination.
“A little bit,” Sen. Kevin Cramer (D) told CNBC on Tuesday when asked if he was concerned that “Squawk Box” could be caught by regulators. “Anything that prevents big players in that industry from getting into the industry, again, I’m always a little skeptical.”
AI companies have some support in Congress
The industry’s call for a narrow antitrust exemption has some supporters.
Rep. George Whitesides, D-Calif., Sen. Bob Latta (R-Ohio), Sen. Adam Schiff (D-Calif.), and Sen. Jim Banks (R-Ind.) have introduced legislation that would create such a carve-out. This will enable AI companies to share information about AI security threats to prevent, investigate, and mitigate risks. The bill would also allow us to work together to reduce AI security issues by slowing or limiting the introduction of high-risk models.
Rep. Sam Licciardo, D-Calif., who sponsored Whiteside’s bill, told CNBC that the bill “needs to be very narrow in scope to focus on exactly what we need to do in terms of testing, evaluation and training to make these models safer.” “Frankly, there’s not a lot of incentive for these companies to collaborate on different things. This industry is very competitive.”
