White House Weighs Tighter Controls on Advanced AI Models

The Trump administration is considering a series of executive actions aimed at addressing national security risks posed by advanced artificial intelligence models, according to industry representatives and policy advisers.

Talks that began in early May 2026 have included the possibility of an executive order that would create a formal review process for “frontier” AI models. Under that approach, AI companies could be required to obtain government approval before releasing certain advanced systems.

The discussions come asourceser the administration reached voluntary agreements with Micrososources, xAI and Google DeepMind for pre-release safety testing intended to evaluate national security risks ahead of model launches.

Beyond a vetting regime, the White House is also weighing provisions that would seek to prevent private-sector “interference” with the government’s use of AI and to tighten contracting and termination standards for federal vendors. The proposals may be influenced by a prior dispute involving Anthropic and military use of its models.

A drasources executive order under consideration is described as 16 pages long and could also include technical guidance for securing open-weight AI models. The deliberations may involve the intelligence community in efforts to safeguard systems against advanced AI threats.

Industry stakeholders have warned that expanded government oversight could slow innovation and technological development. The administration has not announced any final decision.