David Sacks’ White House downfall: Inside the crash and burn

On Monday, The New York Times reported that the White House was weighing federal review of AI models before public release. To the average tech observer, this looked like a complete policy reversal from Donald Trump. For the past year, he had championed industry-friendly deregulation: repealing Biden’s sweeping AI safety executive order, lifting export controls on advanced chips, and signing orders that would penalize states for passing their own AI laws. Now, the administration appeared to demand pre-market federal oversight and vetting.

But inside Washington, this shift stemmed from three major developments. First, Anthropic’s Mythos genuinely rattled the national security apparatus, forcing the administration to confront the risk of adversaries using American AI to attack U. S. public and private sectors. Second, other nations began crafting their own AI regulations, potentially in ways that conflict with American interests. (And yes, “destroying a Big Tech data center in a targeted drone strike” qualifies as a form of government AI regulation, but more on that shortly.)

Third, David Sacks was pushed out as the AI and crypto czar, removing one of Silicon Valley’s direct lines to pitch a pro-industry, “innovation-at-all-costs” agenda to Trump.

Political influence around Donald Trump is notoriously fluid. He takes calls from almost anyone and acts on advice when it suits him. (Remember Laura Loomer’s sway over the National Security Council?) But legally, Sacks,the billionaire venture capitalist and Trump fundraiser,no longer holds the privileges of a special government employee: access to sensitive information, the ability to speak for the White House, or official influence over agencies and staff.

Instead, this “special government employee,” originally meant to serve only 130 days but who lingered for a full year, actively undermined the administration and burned bridges with political allies. Under Sacks, the White House pushed beyond simple deregulation. They tried twice to get Congress to pass a moratorium on state AI laws, and when that failed, attempted an executive order granting the administration power to sue states enforcing such laws. But his Valley-style tactics,along with efforts to consolidate AI policy by sidelining existing agencies,angered Republican and MAGA allies while alienating large parts of Trump’s base. So unsuccessful was this approach that when unnamed White House officials recently pressured certain red states to drop pending AI legislation, claiming it contradicted Trump’s agenda, four GOP state lawmakers went on the record with The Wall Street Journal. (Though if the goal was simply to kill those bills early, it worked.)

Even if Sacks hadn’t self-destructed,and that’s before mentioning his public criticism of Trump, who does not take criticism lightly, over the continued war with Iran,the job itself was becoming too complex for one part-time employee with private-sector ties. In recent months, America’s AI policy scope expanded far beyond Sacks’ 2025 pro-innovation mandate, into areas where a lack of regulation would be wildly irresponsible: national security and geopolitical stability.

A key turning point was the leak of Anthropic’s Mythos, an AI model so adept at finding cybersecurity vulnerabilities that the company,whose reputation depends on acting more responsibly than rivals,refused to release it publicly. The prospect of a Mythos-level model entering the market spooked the national security apparatus and the financial industry, grabbing the attention of three powerful White House figures: Treasury Secretary Scott Bessent, Commerce Secretary Howard Lutnick, and Chief of Staff Susie Wiles.

When Bessent and Wiles met with Anthropic CEO Dario Amodei in April, it signaled they were taking the threat seriously. They also overrode Anthropic’s detractors in the Pentagon, who months earlier had convinced Trump that Anthropic was “woke” and should be banned from government use.

“The national security implications of something like Mythos are hard to deny, and legitimately urgent national security issues are not easy to politicize,” Charlie Bullock, a senior research fellow at the Institute for Law and AI, told The Verge. “Once serious national security people get involved, it’s hard to dismiss or politicize the issue.”

In recent weeks, federal agencies that Sacks had sidelined are gaining authority. On Tuesday, the Commerce Department announced that the Center for AI Standards and Innovation (CAISI) would conduct pre-deployment testing on commercial frontier AI models before release, and had already struck agreements with xAI, Microsoft, and Google DeepMind. CAISI is run by NIST, which was gutted last year by Elon Musk’s Department of Government Efficiency (DOGE) but has begun hiring technical staff.

Other countries are also flexing their muscles in ways the U. S. can’t control. The European Union is debating revisions to the AI Act, and while recent negotiations between EU countries and the European Parliament failed to reach an agreement, whatever emerges will directly impact how frontier AI models are developed,possibly in ways that inadvertently harm America’s business and national security interests.

“[Bessent] really doesn’t like the Europeans,” a tech policy adviser close to the administration told The Verge. The adviser argued that the EU’s proposed privacy regulations would hurt American companies while allowing China to develop faster, citing historical precedent: “We’ve seen this movie before when it came to broadband, where they tried to do the same thing to American broadband companies. In the end, they were just helping Huawei.”

Then there are rogue players who simply don’t care what Sacks or the U. S. government think. Days after U. S. forces bombed Tehran and killed its religious leader, Iran conducted drone strikes on two AWS data centers in the UAE and indirectly damaged a third in Bahrain, causing major power outages across the Middle East and damaging critical infrastructure. Weeks later, Iranian state media announced it would target 18 major U. S. tech companies with regional presence, including Google, Meta, Microsoft, Palantir, and Nvidia, and later claimed to have hit an Oracle data center in the UAE. (UAE media later clarified that an Oracle building in Dubai suffered minor damage from falling debris from an aerial drone interception.)

“While there’s a lot of politics around it here in the United States, where Maine is trying banning them, and DeSantis is talking about banning them, as far as the world is concerned, this is critical infrastructure,” the tech policy adviser told me. “And that’s why one of the first things the Iranians did was bomb not just one, but two of Amazon’s critical data centers, because they know how important it is.” The damage to AWS’s data centers, which serve the entire Middle East, is severe enough that even if the war ended now, it would take “several months” to resume full operations, according to the company.

This doesn’t mean David Sacks has zero influence in the Trump administration. He still has Trump’s direct cell number, and he remains a billionaire CEO,a credential Trump values more than any expertise. The Atlantic’s George Packer recently profiled Sacks, highlighting his plutocratic tenacity. You can’t really stop a master of the universe from flexing occasionally. But even among Trump’s favorite rich guys, Sacks may not rank highly. Last week, Trump hosted a state banquet for King Charles III’s U. S. visit. The guest list included Tim Cook, Jensen Huang, Jeff Bezos, Marc Andreessen, Marc Benioff, and corporate leaders from Meta and Alphabet,but no Sacks, who had attended a previous state banquet at Windsor Castle last year while still at the White House.

When I asked a DC insider familiar with state dinner politics if Sacks had been invited, the answer was telling: “Why would he be? He’s not White House inner circle.”