Trump's AI czar and the governance vacuum

President Trump announced Sunday that Jay Clayton, the former Securities and Exchange Commission chairman whom he elevated last year to Director of National Intelligence, will now lead a new White House body called the “Super Intelligence Force,” coordinating the federal government’s engagement with artificial intelligence. The announcement follows what Trump described as a “Super Intelligence summit” held at the White House last week, details of which were reported by the BBC. Clayton, a capable lawyer whose SEC tenure was marked by a light-touch approach to crypto regulation and a general instinct toward deregulation, has no known technical background in AI. He comes to the role as a loyalist appointee whose primary qualification appears to be the president’s trust. The move crystallizes a pattern: each successive administration has recognized that artificial intelligence is a matter of national strategic importance, and each has responded with an organizational structure that gestures at seriousness without achieving it.

The received wisdom

The progressive-technocratic critique writes itself: Trump is installing a crony with zero relevant expertise to oversee the most consequential technology in human history, while simultaneously gutting the regulatory bodies — the FTC, the NIST AI Safety Institute, the nascent interagency coordination mechanisms — that might actually impose discipline on the industry. The Bank of England’s governor recently warned that the AI boom could trigger market shocks of the kind regulators are poorly equipped to handle. The European Union has its AI Act. China has its own state-directed AI governance infrastructure. The United States, under Trump, has a branding exercise called “Super Intelligence Force” run by a former Wall Street regulator who spent his career facilitating capital markets, not managing transformative technology risk.

This critique is substantially correct. Let’s say so plainly.

A different read

And yet — the right response to a bad governance structure is rarely more governance structure, and the history of Washington’s attempts to institutionalize tech oversight should give pause to those reaching for the regulatory lever. Consider the record: the Cybersecurity and Infrastructure Security Agency, created with bipartisan fanfare in 2018, has struggled to achieve coherent authority over a sprawling threat landscape. The FTC’s various forays into platform regulation have produced years of litigation and little behavioral change. The EU’s General Data Protection Regulation, celebrated as the gold standard of digital governance, created a compliance industry worth billions while doing relatively little to alter the fundamental power asymmetries of the surveillance economy.

The deeper problem with AI governance is not institutional but epistemic. No one — not Clayton, not the most technically sophisticated regulator in Brussels, not the AI safety researchers at Anthropic or DeepMind — can reliably specify in advance what the harms of frontier AI systems will be or at what capability threshold they occur. Regulatory frameworks built on this uncertainty tend to either under-regulate (because the harms haven’t materialized yet) or over-regulate (because political fear runs ahead of technical understanding). The Bank of England’s warning about market shocks from AI is well-intentioned, but the Governor himself acknowledged that regulation is “not the right place to start.” That is a remarkably honest admission from a central banker, and it implicitly endorses a more iterative, less prescriptive approach.

Where the Trump administration’s posture becomes genuinely dangerous is not in its deregulatory instinct — which has intellectual defensibility — but in its apparent indifference to the national security dimensions of AI development. Clayton’s background is in securities law and financial market structure. The relevant domain for an “AI Czar” in 2026 is not Wall Street; it is the intersection of large language models with biological design, autonomous weapons systems, critical infrastructure vulnerability, and the intelligence cycle. China’s AI development is explicitly oriented toward these domains. The absence in the Trump White House of anyone with serious technical and strategic depth in this area is a genuine gap — not because regulation is needed, but because strategic direction is.

The irony is that the private sector has, in some respects, moved faster and more seriously on AI safety than any government. The voluntary commitments extracted from major labs by the Biden administration, whatever their limitations, represented a more substantive engagement with actual risk than a rebranding exercise. Naming a securities lawyer as AI czar while the Pentagon’s own AI safety board has been effectively defunded is less a governance philosophy than a governance abdication.

What to watch

  • Clayton’s first 100 days: Does the “Super Intelligence Force” produce any actual policy outputs — procurement standards, export control guidance, interagency protocols — or remain a communications vehicle?
  • Congress: Are there bipartisan efforts in the lame-duck period to pass even minimal AI safety legislation, following the failure of previous attempts?
  • Chinese capability signals: Beijing has been notably quiet about its frontier AI progress since mid-2026. Any disclosure of new capabilities will test whether Washington has a strategic response framework.
  • Market reaction: The Bank of England’s warning about AI-driven market volatility will look prescient if any large-scale AI-related financial shock materializes before Clayton has a staff.

— J