Trump Administration Eyes Federal AI Licensing as 70% Oppose Local Data Centers

5 articles · Updated · Fortune · May 19

Washington is shifting from dismantling AI rules to considering a federal licensing regime for advanced models, extending a White House rethink that had already included possible pre-release testing.
Two forces are driving the pivot: public opinion has turned sharply against AI, with 70% opposing local data centers and two-thirds saying government has done too little to regulate it.
Anthropic’s Mythos model added urgency by convincing officials that frontier AI carries dangerous dual-use cyber capabilities; the administration has already reportedly blocked parts of its rollout and tightly controlled access.
That security concern is also reshaping foreign policy after Trump’s China trip, with the U.S. now apparently agreeing to AI safety talks with Beijing despite earlier resistance to any such engagement.
The broader result is a rapid collapse in Washington’s anti-regulation stance, with bipartisan support rising on Capitol Hill and some form of AI legislation looking increasingly likely after November.

As AI now finds security flaws faster than humans, how can our nation's cyber defenses realistically keep pace?

Why are federal cybersecurity budgets being cut just as the threat from AI-powered cyberattacks is rapidly escalating?

How can the US regulate powerful AI without stifling the innovation needed to maintain its global technological lead?