Commentary: The White House’s decision to put national intelligence chief Jay Clayton in charge of a new “Super Intelligence Force” is a revealing choice. It treats artificial intelligence not merely as an economic technology or regulatory challenge, but as a strategic contest in which intelligence agencies should help direct the national response. That instinct is understandable. It is also insufficient.
Recent reporting says Clayton will coordinate the government’s engagement with AI, alongside senior officials from the Federal Trade Commission, the Pentagon and the Office of Personnel Management.[NPR] The arrangement could reduce the bureaucratic fragmentation that has long slowed technology policy. But it risks making secrecy and executive control the default language of civilian AI governance.
The case for a central command
The administration has a reasonable argument. Advanced AI systems touch national security, critical infrastructure, labor markets and public information at the same time. A government divided among agencies may miss threats that cross institutional boundaries. Intelligence officials can identify foreign attempts to steal models, manipulate data or exploit vulnerabilities, while defense agencies understand the consequences of increasingly capable systems.
There is also a practical need for speed. Companies are deploying models faster than legislatures are writing rules. Export controls, research funding and cybersecurity protections cannot be managed as unrelated files. A central coordinating office could establish common standards, prevent contradictory regulations and give smaller agencies access to technical expertise they cannot build alone.
Europe, meanwhile, is continuing to pursue an international AI safety framework despite Washington’s opposition to broad global regulation, according to Politico Europe.[Politico Europe] If the United States wants influence over those rules, it needs a coherent position rather than a procession of disconnected initiatives.
Why concentration is dangerous
Coordination, however, is not the same as command. An intelligence-led structure may be well suited to classified threats, but it is poorly designed to resolve public questions about employment, discrimination, privacy and speech. Those issues require evidence that citizens, journalists, courts and elected representatives can examine.
The risk is not simply that officials might make the wrong decision. It is that the public may be unable to determine why a decision was made. A system that treats AI primarily as a national-security asset could encourage secrecy around commercial failures, weaken disclosure requirements and push ordinary regulatory disputes into an emergency framework.
That concern is especially serious when the same administration is considering a committee to oversee the industry, as Politico has reported.[Politico] Oversight bodies must be independent enough to challenge powerful companies and officials alike. They should not become advisory panels whose members are selected for loyalty or access.
What responsible leadership would require
Our view is that the United States should keep a strong national coordinator, but place clear limits around the office. The coordinator should publish regular assessments of AI risks, disclose conflicts of interest, explain major policy decisions and report to Congress in public wherever classification is not genuinely necessary. Independent inspectors general and courts should retain meaningful authority over abuses.
Security agencies should lead on espionage, military applications and infrastructure threats. Civilian regulators should lead on consumer protection, labor, competition and civil rights. The two spheres must cooperate without allowing one to swallow the other.
Industry also deserves a seat at the table, but not the head of it. The technology sector has valuable expertise and enormous incentives to minimize delays, liability and disclosure. Its warnings about regulation can be sincere while still reflecting commercial interests. Public policy should therefore rely on independent testing, academic research and input from workers and affected communities—not only on executives who build the systems.
America does need urgency. It does not need to confuse urgency with unchecked authority. The best AI strategy would protect national security while preserving the democratic habits that make security legitimate: transparency, contestability and accountability. A “Super Intelligence Force” can help organize government. It cannot be allowed to become a substitute for governing in public.