White House Reviews Pre-Approval for AI

The Trump administration is reviewing federal pre-approval before AI models are released, effectively pivoting from its previous pro-industry, anti-regulation stance.

There are three key drivers behind this.

• Anthropic's Mythos has alarmed national security officials with its ability to detect cybersecurity vulnerabilities.
• Growing concern that tightening regulations abroad, such as in the EU, could affect U.S. companies and security interests.
• The White House's AI and crypto czar David Sacks has seen his influence wane, weakening the industry-first agenda.

The Commerce Department's CAISI has taken on pre-deployment testing of commercial frontier AI models, and has already signed agreements with xAI, Microsoft, and Google DeepMind. This shows that Sacks' strategy of blocking state-level AI regulation and concentrating authority at the federal level is losing momentum.

At the same time, as seen in cases like the attack on data centers in Iran, AI infrastructure is becoming a target of national security and geopolitical risk, moving the situation toward a phase where "innovation without regulation" alone is difficult to sustain.