Trump Signs AI Order Imposing 30-Day Model Delay After Bessent Lobbying

3 articles · Updated · The Verge · Jun 10

A June 2 executive order created a government review board for unreleased advanced AI models, allowing officials to delay launches by up to 30 days.
Scott Bessent’s last-minute lobbying helped revive the measure after Trump had backed away from signing it on May 21, when David Sacks and Elon Musk pushed against the original plan.
The first version, reported on May 20, would have allowed delays of as long as 90 days, making the signed order a narrower compromise after competing pressure inside Trump’s orbit.
The episode underscores how AI policy in Washington is being shaped by direct industry access and volatile political calculations ahead of the 2026 midterms.

Sources

Left75%

Center25%

The Verge2h ago

President Trump Signs AI Executive Order with 30-Day Delay After Scott Bessent's Lobbying

CNBC2h ago

Trump signs AI order asking companies to give govt early access

KPLC2h ago

Trump signs an executive order to vet top AI models for national security risks

4 Sources

While the administration pushes deregulation, its own agencies demand stricter AI safety. Which vision for AI's future will ultimately prevail?

As tech giants lobby to dismantle safety measures, can ethical principles ever truly guide the future of artificial intelligence?

With tech spending over $400 million on elections, is public interest being sold to the highest bidder in the race for AI dominance?

U.S. 2026 AI Oversight: Inside Trump’s Voluntary 30-Day Review Order for Advanced Models and Its Geopolitical Stakes

Overview

On June 2, 2026, President Donald Trump signed a landmark Executive Order that established a voluntary framework for government oversight of advanced artificial intelligence models. This order allows the U.S. government to review powerful new AI models for up to 30 days before public release, aiming to ensure that rapid AI advancements are matched with proper safeguards. The immediate driver was growing concern over cybersecurity risks, highlighted by Anthropic’s Mythos model, which exposed software vulnerabilities. By introducing this review system, the Executive Order seeks to prevent advanced AI from being misused or creating new security threats, while encouraging responsible innovation.

...