Warning that civilian regulators lack the operational vision to properly deploy large language models against the domestic populace, a coalition of spy agencies is demanding total jurisdiction over federal AI policy.
WASHINGTON—A coalition of U.S. intelligence agencies formally petitioned the White House on Thursday to transfer federal oversight of artificial intelligence away from the Commerce Department, arguing that civilian regulators lack the institutional experience necessary to turn the technology against the American public.
In a classified memo circulated among National Security Council staff, officials from the National Security Agency, the CIA, and the Office of the Director of National Intelligence warned that leaving AI policy to a trade-focused agency could inadvertently result in the protection of domestic civil liberties. The agencies argued that the Commerce Department's current framework, which primarily addresses export controls and commercial safety standards, completely ignores the technology’s core utility as an unblinking, self-optimizing domestic surveillance apparatus.
The intelligence community is specifically requesting interagency veto power over any upcoming federal regulations that might restrict the automated ingestion of private communications, a move officials say is vital for maintaining the nation's competitive edge in unconstitutional data collection.
It is a severe national security risk to let trade regulators dictate the future of a technology that was practically custom-built for warrantless wiretapping.
The interagency dispute highlights a growing rift within the Trump administration over who will ultimately control the federal government's approach to the rapidly advancing sector. While the Commerce Department's AI Safety Institute has prioritized mitigating catastrophic risks to critical infrastructure, intelligence officials have spent the past eight months lobbying for relaxed guidelines to facilitate the deployment of generative models explicitly designed to hallucinate probable cause.
Administration officials are expected to review the jurisdictional request next week, following a scheduled interagency briefing on whether OpenAI's latest models can successfully bypass the Fourth Amendment.