Overview

The Administration's Theory of Technological Power

The animating logic of the administration's technology policy is that American leadership in artificial intelligence and advanced semiconductors is a strategic asset that the federal government must actively cultivate, protect from foreign appropriation, and liberate from regulatory friction at home. The administration's July 2025 "Winning the AI Race" Action Plan set out more than ninety federal policy actions across three pillars to maintain U.S. global leadership in artificial intelligence. That document established the rhetorical and structural foundation for everything that followed: AI competitiveness framed not as a commercial question but as a national security imperative. The practical consequence has been a consistent preference for executive action over legislation, for industrial partnership over regulatory oversight, and for a strong federal hand against state-level experimentation.

Where Congress has struggled to legislate, the Trump administration has moved aggressively through executive action. That pattern has generated speed and coherence of a kind, but it has also exposed the administration to legal challenge, internal contradiction, and the risks that accompany policymaking that bypasses the deliberative constraints that statute would ordinarily impose.

Artificial Intelligence, Chips, and Industrial Competition

The administration's AI regulatory posture has been explicitly permissive. Beginning with a December 2025 executive order, the administration urged Congress to adopt a federally unified, innovation-oriented regime centered on preemption of state AI laws and a "light-touch" regulatory approach. On March 20, 2026, the White House released its National Policy Framework for Artificial Intelligence, outlining recommendations intended to establish a nationally uniform approach to AI regulation. The framework's seven pillars — covering child protection, infrastructure, intellectual property, free speech, innovation, workforce preparation, and preemption of state laws — reflect a preference for a sector-specific, federally led regulatory model with significant preemption of state AI laws.

The December 2025 executive order directed a national policy to establish a "minimally burdensome" standard for AI, with the Department of Justice empowered to challenge state laws deemed inconsistent with that goal. A proposed ten-year moratorium on state AI laws had already collapsed in a 99-to-1 Senate vote, illustrating the political difficulty of the preemption agenda even within the president's own party. The March 2026 framework advances the argument legislatively, but legislative adoption remains uncertain, even if the proposal is likely to shape near-term congressional activity and longer-term regulatory architecture.

The semiconductor supply chain has produced the administration's most volatile and contested decisions. The second Trump administration has both tightened and loosened export controls on chips, adding dozens of Chinese entities to the Entity List in early 2025 while separately requiring Nvidia to apply for a license to sell its H20 chip to China. The more dramatic shift came later in the year. In August 2025, President Trump announced that Nvidia would receive export licenses to sell its H20 chips to China on the condition that the company pay fifteen percent of its revenues from those sales to the U.S. government; in December 2025 he extended the approach to Nvidia's more advanced H200 chip at a rate of twenty-five percent, with similar arrangements for AMD and Intel. Revenue-for-access is now the administration's de facto export control policy for advanced AI chips. Lawfare has argued that this arrangement is legally unauthorized, because the administration is effectively taxing chip exports without congressional authorization.

The policy logic behind the shift is disputed. Effective export controls must balance the specific advantage the U.S. aims to preserve against potential damage to other interests such as global market dominance, and as CSIS has documented, well-intentioned but poorly conceived export control efforts have often backfired. The effectiveness of controls appears limited in any case, since even with them in place, Chinese firms such as DeepSeek and Alibaba developed highly capable open-source AI models that gained international popularity in 2025. Against that backdrop, Brookings has argued that loosening controls on mature chips may be defensible. Critics from the prior administration have rejected that framing, contending that the arrangement surrenders a durable strategic advantage for near-term revenue.

Cybersecurity and Critical Infrastructure

The administration's cybersecurity posture combines an ambitious strategic vision with deep cuts to the institutional capacity needed to execute it. In June 2025, the administration issued an executive order on strengthening national cybersecurity that kept the existing federal framework intact but redirected it toward priorities including AI, post-quantum cryptography, software supply chain security, and countering foreign actors. A subsequent national cyber strategy released in March 2026 called for AI-powered cybersecurity solutions across federal networks. The administration's stated focus on offense and foreign threat is coherent as far as it goes.

What is harder to reconcile with that focus is the treatment of the Cybersecurity and Infrastructure Security Agency. CISA lost millions in funding and roughly a third of its workforce through buyouts and budget cuts during the first year of Trump's second term. In March 2025, CISA cut nearly half the total budget for the Multi-State Information Sharing and Analysis Center, which provides threat detection services to state and local governments; six months later, it severed ties with the Center for Internet Security, a nonprofit offering similar services. The administration's rationale, stated plainly in budget documents, is that CISA was "more focused on censorship than on protecting the Nation's critical systems." The specific grievance concerns CISA's misinformation work, which the administration has argued disproportionately targeted conservative voices.

The practical effect of those cuts is contested but sobering. Former officials and industry sources worry the reductions have left CISA poorly positioned to respond to AI-fueled hacking threats posed by advanced models. The cuts could weaken state and local capacity, strain relations with industry and allies, and leave gaps in protection against lower-profile but frequent cyberattacks from ransomware groups and state-linked actors. As of mid-2026, CISA's acting director has signaled plans to hire back more than three hundred staff for mission-critical roles, suggesting some recognition that the reductions went too far.

Platform Power, Speech, Privacy, and Surveillance

The administration has not advanced a coherent framework for platform regulation, privacy, or domestic surveillance. Its actions in these areas have been reactive, ideologically selective, and sometimes internally contradictory. The administration's AI framework calls for protecting free speech from algorithmic suppression and preserving state authority to shield children from online exploitation, while simultaneously pushing to strip states of authority to mandate algorithmic transparency or bias mitigation. The framework calls on Congress to preempt state AI laws that "impose undue burdens" while preserving state authority in areas such as traditional police powers to protect children and prevent fraud. Where the line between permissible state authority and preempted regulation falls has not been clearly defined.

The administration has also used the AI preemption agenda to constrain state-level privacy and accountability rules. The December 2025 executive order established an AI Litigation Task Force within the Department of Justice, charged with challenging state AI laws in federal court on grounds including the Dormant Commerce Clause. The directives raise complex constitutional issues likely to result in prolonged litigation and uncertainty for market participants. No federal consumer privacy legislation has moved through Congress, leaving a gap that the administration has not proposed to fill through statute.

Government Partnerships and Private-Sector Dependence

Across AI, chips, and cybersecurity, a structural feature of the administration's approach is its reliance on large private-sector actors to execute national security objectives. The semiconductor revenue arrangement with Nvidia, AMD, and Intel makes the government's export control policy depend on company compliance with conditions that have no clear statutory basis. The AI Action Plan established the government's AI infrastructure ambitions partly through commitments from major cloud and chip companies to expand domestic capacity. CISA's March 2026 cyber strategy envisions AI-powered defense tools sourced from and operated by private vendors.

This public-private model offers speed and scale. It also concentrates leverage in a small number of dominant firms, raises questions about accountability and conflict of interest, and reduces the government's ability to set standards independently. The June 2025 cybersecurity executive order redirected federal cyber priorities toward software supply chain security, a recognition that critical systems are deeply dependent on private vendors — yet the administration's budget cuts have reduced the federal capacity to audit or enforce those supply chain standards. The pattern raises a genuine governance question: whether outsourcing national security functions to commercial partners produces resilience or a more complex form of vulnerability.

The Test of Secure Innovation

Taken together, the administration's technology and cyber policies reflect a coherent theory — maximize American speed and scale in AI and semiconductors, reduce regulatory friction, concentrate authority federally, and confront adversarial powers directly — that has produced a mixed and sometimes self-defeating record. What is established: the administration has moved fast, produced significant regulatory architecture through executive action rather than statute, and shifted the terms of debate on AI governance and chip export controls in ways that will persist. Neither the AI Framework nor pending draft legislation is yet binding, but they broadly signal how the administration and key senators are looking to regulate AI technologies.

What remains disputed: whether the revenue-for-access chip arrangement advances or undermines strategic competition with China; whether dismantling CISA's state-and-local partnerships improves or degrades national cyber resilience; and whether federal preemption of state AI laws produces a genuine innovation environment or simply removes accountability without replacing it. What has not yet been tested: the durability of the preemption framework against constitutional challenge, the adequacy of a reduced CISA in a threat environment that is growing in sophistication, and whether Congress will convert the administration's legislative recommendations into law. There are early signs of a partial course correction at CISA, with the acting director signaling new hiring for mission-critical roles. Whether that adjustment is sufficient, and whether the broader policy architecture produces secure innovation rather than simply rapid but fragile growth, remains an open question.