Access to frontier AI models, the most powerful artificial intelligence systems currently available to developers and businesses, is being shaped by the Trump administration. According to sources familiar with the matter who spoke to CNBC, the White House is taking steps to decide who gets that access, a role that had previously rested almost entirely with the technology companies building these systems.
What a frontier model is
A frontier model sits at the edge of what AI can currently do. These are the systems that major technology companies have spent enormous resources developing, and they are what governments, militaries, businesses, and researchers most want to use. Control over who can reach these tools is, in practice, control over who gets to work at the leading edge of AI capability.
The power moving to Washington
The shift described to CNBC is structural. When a technology company decides who can use its most capable AI, it sets the terms: pricing, usage policies, approved customers. When a government steps into that role, the calculus changes. An administration can prioritize national security interests, restrict access to foreign competitors, or channel capabilities toward preferred partners in ways that a private company cannot do unilaterally.
The sources did not describe specific policies, timelines, or which companies are most directly affected. What they conveyed to CNBC is directional: the White House is asserting authority over the access layer, pulling that decision away from tech giants.
What the reporting does not yet confirm
The source material does not specify which frontier models fall under this authority, which federal agencies are involved, or how any restrictions would be enforced. Those details matter for companies and researchers who depend on these systems. The reporting reflects early-stage sourcing, and no formal mechanisms have been publicly described.