The Role of Washington, DC in AI Governance in the United States: Tensions Between Federal and State Law

Introduction

The US remains the global leader in the development and deployment of artificial intelligence (AI), with the AI sector representing a rapidly growing and increasingly influential part of the broader domestic economy. The US also continues its guiding role in policy debates over AI governance. The centre of that debate is Washington, DC, the nation’s capital and policymaking nucleus of the US government. As the White House, federal departments and agencies, and Congress continue to shape the contours of national (and international) AI policy, developments in Washington, DC will continue to influence global standards and governance approaches too.

The current Trump Administration has adopted a strong innovation and markets orientation to technology policy writ large, generally favouring more limited regulation. More specifically to AI, President Trump’s July 2025 America’s AI Action Plan states in the introduction: “The United States needs to innovate faster and more comprehensively than our competitors in the development and distribution of new AI technology across every field, and dismantle unnecessary regulatory barriers that hinder the private sector in doing so.” The White House has repeatedly called upon Congress to pass a pre-emptive federal AI law to limit patchwork regulation among the states, and several bills have been introduced – although none have reached the President’s desk for signature. On 20 March 2026, the Trump Administration also issued the National Policy Framework for Artificial Intelligence, non-regulatory recommendations consisting of seven pillars:

• 1. Protecting Children and Empowering Parents;
• 2. Safeguarding and Strengthening American Communities;
• 3. Respecting Intellectual Property Rights and Supporting Creators;
• 4. Preventing Censorship and Protecting Free Speech;
• 5. Enabling Innovation and Ensuring American AI Dominance;
• 6. Educating Americans and Developing an AI-Ready Workforce; and
• 7. Establishing a Federal Policy Framework, Preempting Cumbersome State AI Laws.

America’s AI Action Plan and National Policy Framework for Artificial Intelligence are discussed in more detail below.

There is also a robust body of existing consumer protection and sector-specific laws and regulations that form the basis of a mature US AI legal regime. However, the Trump Administration is closely scrutinising emerging state-based AI laws to ensure they do not unduly burden innovations in AI, sustaining US dominance in the field.

While the bulk of this article focuses on US federal governance, at the local level, the District of Columbia’s AI regulatory policy prioritises risk management and accountability with respect to the local DC government’s AI use. The Council of the District of Columbia, the District’s legislative body, has previously proposed legislation addressing algorithmic transparency and automated decision-making (D.C. Bill 24-558), echoing the District’s well-known consumer protection and privacy legal regime. At the District level, Washington, DC has taken a cautious, governance-first approach to AI, focusing on public sector use, transparency, and the application of existing consumer protection and civil rights laws, while continuing to explore more comprehensive legislation. However, Congress has ultimate authority over local DC laws, and can review and overturn such laws. And, as articulated in the President’s 11 December 2025 Executive Order 14365, wide-reaching state legislation may come into conflict with a potential federal AI framework, as well as from the US DOJ as discussed further below.

Legal developments from the White House

As noted, the Trump Administration has favoured a deregulatory approach for the AI industry, advocating for a uniform, minimally burdensome national framework and taking a critical stance toward the growing number of state-level AI laws. Despite this, the absence of a comprehensive federal AI law means that state laws – even if challenged by the Trump Administration as contemplated in Executive Order 14365 (“Ensuring a National Policy Framework for Artificial Intelligence”) – will continue to serve in the interim as a dominant policy and compliance model. Nevertheless, over the past year the authors saw notable developments in the federal AI landscape in the form of Executive Orders, targeted enforcement decisions, and issue-specific statutes.

Executive Orders and other activity

Since January 2025, the Trump Administration has taken several affirmative steps to advance the White House’s AI policy, including the above-referenced America’s AI Action Plan, Executive Order 14365 (“Ensuring a National Policy Framework for Artificial Intelligence”), and the March 2026 National Policy Framework for Artificial Intelligence. These initiatives are not legislation, but they do collectively outline the Administration’s approach to AI governance and more generally the domestic AI industry. In particular, they emphasise a co-ordinated federal framework in place of fragmented state-level regulation, identify AI innovation and US leadership as central to economic strength and national security, and call for safeguards to ensure that AI systems are secure, reliable, and not based on ideological bias. A more comprehensive list of White House AI policy includes the following.

• Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence (23 January 2025):
1. declared that it is the policy of the United States “to sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security” (Exec. Order No. 14179, Section 2);
2. revoked Executive Order 14110, Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence (30 October 2023) (See id. Section 5); and
3. initiated the development of America’s AI Action Plan (See id. Section 4).
• Executive Order 14319, Preventing Woke AI in the Federal Government (23 July 2025):
1. declared that “it is the policy of the United States to promote the innovation and use of trustworthy AI” developed in accordance with “truth-seeking” and “ideological neutrality,” referred to jointly as “Unbiased AI Principles.” (Exec. Order No. 14319, Section 3.)
• America’s AI Action Plan (23 July 2025):
1. called for increased federal investment in AI research and development, fostering public–private partnerships to develop “full-stack AI export packages,” and supporting the commercialisation of cutting-edge AI technologies (see America’s AI Action Plan, at 4–5, 12, and 20);
2. called for establishing the physical, digital, and human capital foundations necessary for robust AI development and deployment, specifically through expanding data centre capacity and enhancing the power grid (see id. at 14–19);
3. tasked federal agencies overseeing AI-related discretionary funding to ensure that federal resources are not allocated to states with regimes deemed to hinder the effectiveness of such funding (see id. at 3); and
4. generally promoted the export of America’s full AI technology stack internationally, while also encouraging “partners and allies to follow U.S. controls” (Id. at 21–22).
• Executive Order 14365, Ensuring a National Policy Framework for Artificial Intelligence (11 December 2025):
1. directed the Attorney General to establish an AI Litigation Task Force with the “sole responsibility ... to challenge State AI laws inconsistent with the policy set forth in section 2 of this order, including on grounds that such laws unconstitutionally regulate interstate commerce, are preempted by existing Federal regulations, or are otherwise unlawful in the Attorney General’s judgment, including, if appropriate, those [state AI] laws identified pursuant to section 4 of this order” (Exec. Order No. 14365, Section 3);
2. directed the Secretary of Commerce, in consultation with the Special Advisor for AI and Crypto, to issue a Policy Notice “specifying the conditions under which States may be eligible for [certain] funding under the Broadband Equity Access and Deployment (BEAD) Program” (Id. Section 5);
3. directed the Federal Trade Commission (FTC) to “issue a policy statement on the application of the [FTC] Act’s prohibition on unfair and deceptive acts or practices under 15 U.S.C. 45 to AI models” (Id. Section 7); and
4. directed the Special Advisor for AI and Crypto and the Assistant to the President for Science and Technology to “jointly prepare a legislative recommendation establishing a uniform Federal policy framework for AI that preempts State AI laws that conflict with [Federal] policy,” noting that the recommendation shall not propose pre-empting state AI laws relating to: child safety protections, AI computing and data centre infrastructure, and state government procurement and use of AI (Id. Section 8).
• National Policy Framework for Artificial Intelligence (20 March 2026):
1. urged Congress to adopt a national AI policy framework designed to protect children, safeguard and strengthen American communities, respect intellectual property rights, prevent censorship, support innovation, develop an AI-ready workforce, and prevent a fragmented patchwork of state regulations (see National Policy Framework for Artificial Intelligence, Sections I–VII);
2. recommended that Congress establish enablers for innovation, such as regulatory sandboxes and resources to make federal datasets accessible to industry and academia in AI-ready formats for use in training AI models and systems (see id. Section V);
3. advocated for federal pre-emption of state AI laws that impose “undue burdens,” while preserving certain areas of state authority, including generally applicable laws protecting children, prohibiting fraud, and protecting consumers, as well as state zoning authority over AI infrastructure and rules governing a state’s own procurement and use of AI (see id. Section VII); and
4. aimed to ensure appropriate AI infrastructure development that also protects local communities from harmful impacts (see id. Section II).

A case study on federal policy in action: the FTC’s setting aside of the Rytr Order

Eleven days after the White House issued Executive Order 14365 (“Ensuring a National Policy Framework for Artificial Intelligence”), the FTC re-opened and set aside a final consent order against Rytr, an AI-enabled content generator and writing assistance service, noting in the accompanying press release that “the complaint [filed during the Biden Administration] failed to satisfy the legal requirements of the FTC Act and that the order unduly burdens artificial intelligence (AI) innovation in violation of the Trump Administration’s Artificial Intelligence Executive Order and America’s AI Action Plan.” (See FTC, FTC Reopens and Sets Aside Rytr Final Order in Response to the Trump Administration’s AI Action Plan (22 December 2025).)

By way of background, in September 2024, the FTC brought an enforcement action against Rytr, alleging that the company’s “Testimonial & Review” feature allowed subscribers to generate false and deceptive online reviews, including reviews containing specific, often material details that had no relation to the user’s input. The Biden era FTC’s final consent order prohibited Rytr from engaging in similar illegal conduct in the future and barred the company from advertising, promoting, marketing, or selling any service dedicated to – or promoted as – generating consumer reviews or testimonials.

The December 2025 reversal represented the first concrete action taken by the FTC in response to the AI Action Plan’s recommendation that the agency review prior matters that may unduly burden AI innovation. More broadly, the decision quoting Chairman Ferguson’s dissent in the 2024 enforcement action suggested that, while the FTC will continue to pursue AI-related fraud and deception, the Commission’s current stance is that “[t]reating as categorically illegal a generative AI tool merely because of the possibility that someone might use it for fraud ... threatens to turn honest innovators into lawbreakers and risks strangling a potentially revolutionary technology in its cradle.” (See Order Reopening and Setting Aside Order at 5, In re Rytr LLC, Docket No. C-4806 (22 December 2025).)

Nevertheless, the FTC has remained active in more traditional consumer protection matters involving AI-related claims, a trend that began with the announcement of its Crackdown on Deceptive AI Claims and Schemes also known as Operation AI Comply in September 2024. To that point, the FTC sued Air AI on 25 August 2025 over allegedly deceptive claims concerning business growth, earnings potential, and refund guarantees, and announced a proposed settlement on 24 March 2026 that would ban the company from marketing business opportunities. Similarly, on 10 September 2025, the FTC also launched an inquiry into AI companion chatbots under its Section 6(b) investigative authority focused on safety, advertising, and data-handling practices, particularly where children and teens may be affected. And by way of additional example, the FTC entered a final order on 28 August 2025 against Workado, a technology company that provides tools for digital marketing agencies and website owners, relating to allegations that the company’s claims about the accuracy of its AI content detector were unsupported.

Federal AI legislation: the Tools to Address Known Exploitation by Immobilising Technological Deepfakes On Websites and Networks Act – TAKE IT DOWN Act

While Congress has deliberated on the passage of a comprehensive national AI law, only one AI-specific bill was signed into law in 2025, the TAKE IT DOWN Act. The law targets the knowing publication of “intimate visual depictions” of minors and non-consenting adults, including through AI-generated media. The law mirrors, in many respects, the broader legislative trend at the state level, where a majority of states have passed laws to curtail sexually explicit deepfakes and other forms of non-consensual intimate imagery.

The TAKE IT DOWN Act passed with bipartisan support, and President Trump signed it into law on 19 May 2025. Both he and the First Lady championed the law. The TAKE IT DOWN Act imposes criminal liability for the non-consensual publication of intimate images, including AI-generated deepfakes, and requires covered online platforms to establish notice-and-removal systems by 19 May 2026. Under the TAKE IT DOWN Act, covered platforms must remove flagged content within 48 hours of receiving a valid removal notice. (47 U.S.C., Section 223a(a)(3).) Platforms that fail to comply will be treated as having violated the FTC Act and will be subject to FTC enforcement. (47 U.S.C., Section 223a(b).)

National Institute of Standards and Technology (NIST)

The NIST, a component of the US Department of Commerce, is a key government institution impacting the development of AI law and policy. Headquartered in a suburb of Washington, DC, NIST has continued to build on its globally recognised AI Risk Management Framework. Among the nearly 100 policy positions outlined in President Trump’s America’s AI Action Plan was a directive for NIST to revise the framework and remove references to misinformation, diversity, equity, and inclusion, and climate change. This direction reflects an endorsement of NIST’s technical leadership and its role in advancing US influence in global AI standards, as well as alignment with the Administration’s view that AI systems should adhere to principles of “truth-seeking” and “ideological neutrality.” Apart from these anticipated revisions, the NIST AI Risk Management Framework continues to serve as a foundational tool for AI governance across federal agencies, government contractors, and the broader technology industry. The NIST AI Risk Management Framework is also referenced in certain state laws, including Texas’ Responsible Artificial Intelligence Governance Act, in the context of safe harbour provisions and affirmative defences.

Additionally, NIST has continued to advance its cybersecurity guidance and recommendations for the deployment of artificial intelligence technologies. In December 2025, it released a preliminary draft of the Cybersecurity Framework Profile for Artificial Intelligence (Cyber AI Profile), which addresses securing AI system components, enabling AI-driven cyber defence, and managing cybersecurity risks associated with AI adoption. NIST convened stakeholders for a January 2026 workshop to discuss the draft, and NIST published findings from that workshop in a March 2026 blog post. The post discussed a few key themes that emerged during the workshop, including but not limited to, special considerations raised by agentic AI tools, approaches to addressing AI adoption challenges grounded in cybersecurity, and the importance of robust human-in-the-loop processes and training. A month later, NIST issued a Concept Note for an AI Risk Management Framework Profile focused on Trustworthy AI in Critical Infrastructure, intended to guide critical infrastructure operators in identifying and implementing appropriate risk management practices when deploying AI-enabled capabilities. The Concept Note is also designed to help guide critical infrastructure operators towards specific risk management practices to consider when engaging AI-enabled capabilities in a clear and actionable manner.

District of Columbia developments

The District of Columbia’s local approach to AI governance diverges somewhat from the federal government. While the federal government has moved in a more market-friendly direction favouring limited regulation to reduce perceived impediments to rapid AI development and deployment, the District of Columbia has adopted regulations that impose certain principle-driven obligations for local agencies to consider before deploying AI tools. Specifically, in February 2024, Mayor Muriel Bowser issued Order 2024-028, Articulating DC’s Artificial Intelligence Values and Establishing Artificial Intelligence Strategic Benchmarks (“the Order”), which identified six values (“AI Values”) that District of Columbia agencies must consider and document before deploying AI tools. These core principles include:

• 1. clear benefit to residents;
• 2. safety and equity;
• 3. accountability;
• 4. transparency;
• 5. sustainability; and
• 6. privacy and cybersecurity.

The Order also established the AI Advisory Group and AI Task Force, charged with enhancing the District of Columbia’s internal AI governance posture as well as conducting regular community and stakeholder engagement on the AI Values.

Building on the Order, on 12 February 2026, the District became the first major US city to require mandatory responsible AI training for all District of Columbia government employees and contractors, further underscoring its precautionary and risk-conscious approach to AI governance. However, the Stop Discrimination by Algorithms Act (Bill 24-558), the District’s most ambitious AI measure as of publication, never passed. Introduced in 2021 and reintroduced during the 2023–2024 legislative session, the Stop Discrimination by Algorithms Act would have addressed discrimination in algorithmic decision-making in employment, housing, credit, and insurance.

Separate from legislative developments, the Office of Attorney General for the District of Columbia has historically played an active role in addressing issues posed by emerging technologies through its consumer protection authority under the Consumer Protection Procedures Act, D.C. Code, Section 28-3901 et seq. With that said, as of publication, there is no record of Attorney General Brian Schwalb having brought any enforcement actions under that statute based specifically on AI-related activity. However, the District of Columbia has joined bipartisan coalitions of state attorneys general seeking to address certain AI-related harms, including the generation of non-consensual sexually explicit images.

Conclusion

The legal landscape for AI in the United States remains defined less by a settled consensus than by a continuing contest over who should govern AI and on what terms. The federal government has repeatedly championed national uniformity, industrial expansion, and innovation-led policy, yet in the absence of comprehensive pre-emptive federal legislation, states continue to shape the practical compliance environment. Local developments in Washington, DC further illustrate that these tensions are playing out not only between the federal government and the states, but also within the nation’s capital itself, where the District has developed a more precautionary and governance-focused approach to AI regulation. Meanwhile, targeted federal interventions, whether through executive action, agency enforcement, export controls, or issue-specific legislation, are beginning to establish clearer expectations in discrete areas of risk, while encouraging the rapid industry growth needed to maintain America’s dominance and leadership on AI.

For businesses, the central challenge is how to navigate a framework that is simultaneously fragmented, fast-moving, and deeply tied to broader concerns around competition, security, infrastructure, privacy, and consumer protection. That challenge is likely to intensify as the technology itself evolves, particularly with the emergence of increasingly autonomous and agentic systems that do not fit neatly within existing legal categories. In this environment, businesses that approach AI governance as a strategic and ongoing discipline rather than a narrow compliance exercise will be best positioned to manage risk and adapt to legal changes. In this evolving landscape, Washington, DC will remain not only the seat of American government, but also a focal point where competing visions of AI regulation are actively shaped, contested, and translated into policy.