About This Series

The ML Strategies 2026 Policy Outlook Series explores the policy, political, and regulatory dynamics shaping key sectors in the year ahead. Across six installments, our team analyzes how federal action, election-year pressures, and agency decision-making are converging to influence business strategy, investment decisions, and risk management in an uncertain environment.

AI Regulation in 2026: Competitiveness, National Security, and a Fragmented Policy Landscape

Policymakers have acknowledged the risks and promise of artificial intelligence (AI), and 2026 is shaping up to be a year of governing it — without agreement on who should be in charge. AI policy in 2026 is converging around competitiveness and national security. The result is a year likely to be characterized by executive action, agency assertion, targeted legislation, and aggressive state experimentation.

A National AI Framework and the Debate Over Federal Preemption

Momentum continues to build in Congress for a national AI framework, driven primarily by concern over an increasingly fragmented landscape of state laws. Lawmakers worry that divergent state requirements could impose significant compliance burdens on companies developing and deploying AI at scale — and, in the administration’s view, weaken US competitiveness.

Several Republicans, including Sen. Ted Cruz (R-TX) and Rep. Jay Obernolte (R-CA) (chairman of the House AI Task Force) are positioning themselves at the center of this debate and are working on legislation to preempt or limit state regulation and/or establish federal baselines. There is also a proposal to codify President Trump’s executive actions on AI. Taken together, these efforts underscore a broader reality: while lawmakers agree on the need for a national framework, they remain deeply divided over its scope, enforcement mechanisms, and the extent of federal preemption.

Executive Orders and Agency Action: Federal AI Preemption Without New Legislation

Where Congress has struggled to legislate, the Trump administration has moved aggressively through executive action. The December 2025 Executive Order, “Ensuring a National Policy for Artificial Intelligence,” makes federal preemption an explicit objective, declaring that U.S. AI dominance depends on a “minimally burdensome national policy framework” and warning that “excessive state regulation thwarts this imperative,” which ML Strategies previously covered in our newsletter AI: The Washington Report.

That order follows a series of failed legislative attempts in 2025 — including a proposed 10-year moratorium on state AI laws that collapsed in a 99-1 Senate vote — highlighting just how contested preemption remains, even within the GOP.

In 2026, the administration is expected to operationalize its strategy through agencies rather than statutes. More than 90 federal actions outlined in the White House AI Action Plan are moving forward, including the creation of a DOJ-led AI Litigation Task Force. That task force is positioned to challenge state AI laws on grounds ranging from interference with interstate commerce to preemption by existing federal statutes or agency regulations. Further analysis on the July AI Action Plan and the December Executive Order can be found in our newsletter AI: The Washington Report.

The result is not clean preemption, but managed conflict. States retain authority in theory, while federal agencies increasingly assert that existing law is sufficient to govern AI-related harms.

State AI Laws vs. Federal Policy: Compliance Risks in a Patchwork Regulatory System

Meanwhile, state governments will continue to act as AI policy laboratories, enacting detailed compliance regimes that companies operating nationally must navigate. Several of these laws are already taking effect or expanding in 2026, including in California, Colorado, Texas, New York, and Florida. California’s SB 53 is widely viewed as a de facto standard-setter, as described in our newsletter, demonstrating how state laws can shape national compliance norms even in the absence of federal legislation.

This dynamic ensures that preemption will remain one of the most contentious issues on Capitol Hill. It is no longer a clean partisan divide, with Republican-led states like Florida and Texas advancing their own AI frameworks even as GOP lawmakers debate federal overrides. As a result, preemption language may surface in committee drafts and executive actions, but a definitive resolution is unlikely in 2026.

AI Chips, China, and US National Security: Export Controls and Trade Policy

AI governance in 2026 cannot be separated from semiconductor policy.

In a rare moment of near-unanimous bipartisanship, the House Foreign Affairs Committee advanced Chairman Brian Mast’s (R-FL) AI OVERWATCH Act on January 21, signaling growing congressional urgency around AI chip exports to foreign adversaries. The bill would treat advanced semiconductor exports similarly to weapons sales and prohibit the sale of Nvidia’s Blackwell chips to foreign entities of concern (China, Iran, North Korea, Russia, and Venezuela) for two years. That temporary restriction is designed to give U.S. manufacturers time to produce more advanced AI chips domestically, allowing exports of Blackwells to China only after they are no longer at the technological frontier. Despite pushback from industry voices, online influencers, and some White House officials, congressional resistance to AI chip exports has hardened, positioning the issue as a top legislative priority in 2026.

That congressional momentum is unfolding alongsideexecutive action. On January 14, President Trump issued a presidential proclamation adjusting imports of semiconductors, marking a narrowly targeted but consequential shift in policy. Acting under a completed Section 232 investigation, the administration imposed a 25% tariff on a limited set of advanced AI chips imported into the U.S.s. While formally justified as a national security measure addressing reliance on foreign-produced semiconductors, the move is widely viewed as an indirect mechanism to enforce the administration’s earlier demand that the U.S. receive a 25% share of revenues associated with exports of these chips to China, a requirement that could not be imposed directly as an export tax under U.S. law. Critics argue the approach stretches national security authorities without fully resolving underlying domestic manufacturing constraints. The administration has pointed to ongoing trade negotiations and new investment commitments as incremental steps toward strengthening US semiconductor capacity.

Meanwhile, the Department of War (DoW) can be expected to maintain a strong focus on defense acquisition reform, prioritizing speed, flexibility, and performance as it streamlines contracting, expands competition, and improves supply chain visibility. With warfighter readiness as the organizing principle, DoW will continue to strengthen the domestic industrial base through targeted use of multiyear procurement, demand signaling, and strategic stockpiling of critical inputs — including semiconductors and key minerals — to reduce foreign dependence and sustain operational readiness.