Trump administration enforces AI regulation through ad hoc interventions despite anti-regulatory stance

The White House Is Making Up Its Rules for AI in Real Time

It's been nearly a week since the Trump administration sent an export control directive to Anthropic, forcing one of the world's leading AI labs to pull its most advanced models offline. After days of negotiations between Anthropic and the White House, the two still remain at odds about how to bring Claude Mythos and Fable 5 back. Why? Well, it depends whom you ask. Throughout this entire debacle, Anthropic doesn't believe it violated any concrete procedures or rules laid out by the Trump administration, according to a person close to the company. But the White House contends that Anthropic behaved recklessly, demonstrating that it can't be trusted to safely roll out frontier technology. This saga proved to me that we're now officially in the Wild West era of American AI regulation. While there are few laws on the books governing frontier AI development, that doesn't mean companies won't end up in trouble with Trump's White House when they cross unspoken lines. "The problem here is that the White House has been in this extreme anti-regulatory posture, and they're now faced with the real AI capabilities that people have been predicting for many years," says a former White House technology official, who asked to remain anonymous to avoid jeopardizing their professional relationships. "There should have been preparation and policies to systematically deal with this, managing the benefits and risks, but instead it's just this slap-dash approach that puts the AI industry in a real quandary." The Trump administration has repeatedly blocked efforts to impose guardrails on the AI industry, often arguing the rules could hamper US innovation and lead the country to fall behind rivals like China. Since he returned to the White House, President Trump has signed executive orders that reversed a Biden-era effort to create a national AI framework and created a federal task force to challenge state laws that could be deemed onerous. While WIRED and other publications have made details of the negotiations between Anthropic and the White House public over the last week, the dispute is still defined by its opaqueness. At no point has the US government clearly stated what Anthropic did wrong -- the best we have is a post on X outlining the general situation from White House technology adviser David Sacks. Ironically, the White House's actions have likely hampered the very kind of innovation it wants to protect. The Trump administration demanded that Anthropic prohibit all foreign nationals from accessing Mythos and Fable 5, preventing many of the AI lab's own employees from accessing its most cutting-edge models, which the company says have sped up its research and development in recent months. All of Anthropic's customers are locked out, too, including Apple, Meta, and much of the Fortune 500. The White House may have had legitimate reasons to be concerned about Anthropic's models. As my colleagues and I reported on Wednesday, US officials grew concerned when they learned earlier this month that Anthropic shared Mythos with SK Telecom, a South Korean telecom giant they allege has ties to China. Separately, Amazon CEO Andy Jassy raised concerns to US Treasury secretary Scott Bessent that some of the guardrails on Claude Fable 5, a safeguarded version of Mythos, could be circumvented. Even if these worries are legitimate, that doesn't mean the White House handled them adequately. In the first case, Anthropic says it coordinated with the US government on the rollout of Mythos, suggesting there may have been a chance for officials to raise alarm about SK Telecom ahead of time. Anthropic has also been working with the Korean company for years, and the arrangement seemingly never caused it to run into national security problems before. And when the White House raised concerns about SK Telecom to Anthropic, it revoked access to the model immediately, we reported. In the second case, WIRED and other outlets have written at length about how every large language model can be jailbroken to varying extents. Anthropic and independent cybersecurity researchers argue that solving jailbreaks is not an easy or isolated issue. Because AI models are probabilistic, not deterministic, tech companies can't guarantee what exactly they will generate in response to a given prompt. If the White House won't let Anthropic release Claude Fable 5 until the jailbreaking problem is solved, I wouldn't hold my breath. But ultimately, the problem here is not that the US government is trying to ensure that advanced AI models have proper safeguards or stay out of the hands of America's foes. It's that the Trump administration is now in a position where it has to make these regulatory decisions in real time. I can assure you that other AI labs like OpenAI, Google, and Meta have been watching Anthropic's misfortune with bated breath. I've been asking AI executives the same question all week: How can you avoid a fate similar to Anthropic's? So far, many AI leaders are coming to the same conclusion: They believe they will need to give the White House early access to their latest AI models. Additionally, they say, they need to be extremely proactive about sharing information with the Trump administration about upcoming model launches -- the risk of catching officials off guard is simply too great. "Advance notice, advance access. I think those are the primary asks that we've heard, not just from the US, but others around the world," Aidan Gomez, CEO of the Canadian AI lab Cohere, said in an interview earlier this week. "I think those are good things in many respects. It shows strong engagement and consideration by authorities on a super important technology." President Trump signed an executive order last month that created a "voluntary" system for AI labs to submit models to the government for early testing. It included a carve-out explicitly stating that this would not become a mandatory licensing regime, a major concern raised by the tech industry that had held up the executive order in the first place. But after what happened with Anthropic, it seems like the Trump administration has effectively created an ad hoc version of such a regime. "The Trump administration, frankly, should not have said that this was a voluntary regime," says the former White House technology official. "It seems very clear that what they are now doing is a licensing regime." This is an edition of Maxwell Zeff's Model Behavior newsletter. Read previous newsletters here.

Inside the White House's AI power center

Why it matters: Departures among key White House officials, combined with rapid advances in technology, are shaking up who's taking the lead on AI policy in the administration. The big picture: Silicon Valley figures David Sacks and Siriam Krishnan have served as key architects of the administration's AI agenda. But with Sacks stepping back from day-to-day involvement and Krishnan preparing to leave, influence is shifting inside the White House to a broader group of officials and aides. Here's who's running the show -- for now. Commerce Secretary Howard Lutnick. Lutnick's signature was on the letter that sparked the latest confrontation between Anthropic and the administration, ultimately leading to the takedown of the company's Fable and Mythos models. * Last week, Lutnick imposed export controls on Anthropic, effectively creating a licensing regime that could eventually impact other AI labs. * He is now leading meetings on the sidelines of the G7 to discuss expanding access to advanced AI models and standing next to President Trump during his summit press conference. * While Lutnick brokers abroad, Chris Fall at Commerce's Center for AI Standards and Innovation has been holding technical meetings in D.C. Zoom in: Lutnick was once said to be on the outs for going off message in TV interviews and not having a solid handle over his department's jurisdiction, sources familiar with the matter said. * But last week's Anthropic fallout put him back in the game and it's now Lutnick's moment. * "He's fixing a problem. He's not being a problem. And he's doing a great job," said one senior administration official following the letter imposing export controls. Treasury Secretary Scott Bessent. Bessent does not directly oversee AI testing. Yet he was the point of contact when Amazon raised concerns about Anthropic safety issues and was among the few cabinet members flanking the president during a G7 press conference. * Bessent is viewed as "the more reasonable actor" compared to Defense Secretary Pete Hegseth, who has been public about his disdain for the company, said one source familiar with the administration's thinking. * "Bessent is trusted by the private sector and critical infrastructure operators as a sober actor," another source familiar said. White House Chief of Staff Susie Wiles. Wiles, a veteran political strategist, has not typically been involved in day-to-day AI policy. * But she was receptive to Bessent's concerns about how Anthropic's Mythos could impact the financial sector, helping to reopen lines of communication with the company. National Economic Council's Ryan Baasch. Baasch is said to be the person carrying the torch inside the White House for Sacks and Krishnan. * Sacks left earlier this year to join the President's Council of Advisors on Science and Technology and Krishnan plans to leave by the end of this month. * Baasch has worked alongside the two advisers to influence AI policy on Capitol Hill and to push federal preemption of state AI laws. Tension point: National Cyber Director Sean Cairncross is also holding technical meetings in Washington, but a clash with Bessent over the administration's AI response has deepened following the latest Anthropic dispute, sources familiar tell Axios. * Cairncross believes that Treasury has become too involved, while Bessent and allies in the White House believe Cairncross has not met the moment with the necessary urgency. * Cairncross' head of policy and senior adviser Thomas Lind plans to leave, further depleting the Trump administration of technical expertise. * The Office of the National Cyber Director did not respond to a request for comment. What they're saying: "The President's team, including Secretary Bessent and Sean Cairncross, is working closely together to strengthen America's cyber and national security, protect critical infrastructure, and ensure the United States remains the global leader in AI innovation," White House spokesperson Liz Huston said. The bottom line: If personnel is policy, look to this latest cast of characters shaping a U.S. regulatory regime with global reach. Marc Caputo contributed to this story.

Trump's shadow AI policy

Why it matters: The White House says AI shouldn't be regulated, but it's shaping the industry through case-by-case interventions without clear rules, creating major uncertainty. * What has emerged is a shadow AI policy, one that shapes the AI industry's future without ever spelling out the rules. The big picture: The Trump administration has made opposition to AI regulation a cornerstone of its AI agenda, rescinding Biden-era requirements and arguing that excessive rules would slow innovation. * Yet its light-touch, hands-off vision for AI has given way to an ad hoc system of company-specific interventions, voluntary frameworks and executive actions. * Unlike traditional regulation, this influence operates outside of the formal rulemaking process, with few published standards and limited guidance for companies to help navigate the administration's expectations. Friction point: It is supposed to be the role of Congress to make laws the administration then enforces. * But despite recent efforts, including a bipartisan AI safety bill introduced in the House, Capitol Hill is frozen on AI as midterms loom, and the administration is taking the lead with executive action. Without clear national AI rules set by Congress, the administration has focused on overriding state AI laws, the national security and cybersecurity implications of advanced AI models, the procurement of AI systems into the federal government and the economic impact of the biggest AI companies. * Export controls, voluntary testing frameworks, and procurement guidelines are becoming the building blocks of the administration's shadow AI policy and a guide for how AI companies operate in the U.S. * AI CEOs talked with Trum at the G7 summit on Wednesday about the possibility of what OpenAI's Chris Lehane described as a global forum for AI standards. As Anthropic and the administration hash out whether export controls can be lifted on its latest models, other leading AI labs are working out how to comply Trump's latest executive order, which established a voluntary framework for government review of some advanced AI models. * The Anthropic situation also illustrates a central concern for the industry: uncertainty. Without clear rules, companies can find themselves navigating personalities and broader politics as much as policy. * The General Services Administration is also considering a new rule around safeguarding data when LLMs process government information that would set certain privacy and security standards for companies wishing to contract with the government. Zoom out: The administration's moves are not regulation in the sense of Europe's AI Act or the UK's Online Safety Act. But because the U.S. is the home of the world's most advanced models, its decisions matter the most. * At the G7 summit this week, the notion that other countries need to establish "tech sovereignty" so as to not rely on American companies loomed large. * But foreign leaders also know they can't ignore the world's leading models and only rely on homegrown AI. The bottom line: The Trump administration may not call it AI regulation, but its decisions are dictating the future of the technology around the world.