White House Publishes National Policy Framework for Artificial Intelligence
Nick Palmieri
On March 20, 2026, the White House released a National Policy Framework for Artificial Intelligence. The framework sets forth legislative recommendations for seven policy areas: (i) protecting children, (ii) safeguarding communities, (iii) respecting intellectual property rights, (iv) preventing censorship, (v) enabling innovation, (vi) workforce development, and (viii) preempting state AI regulation. The emphasize revolves around the need to protect American creators, publishers, and innovators from AI-generated outputs that infringe their protected content, while simultaneously preserving lawful innovation and free expression.

Though the framework only includes recommendations (that is, it is not a binding rule on Congress or any authorities), it suggests some notable positions with respect to intellectual property, in particular the use of copyrighted materials by AI: while the Administration believes that training AI models on copyrighted material (currently) does not violate copyright laws, it acknowledges the existence of contrary arguments and supports allowing the courts to resolve this question, recommending that Congress refrain from any action that would affect the judiciary's determination of whether such training constitutes fair use. Rather, Congress is encouraged to consider enabling licensing frameworks or collective rights systems that would allow rights holders to collectively negotiate compensation from AI providers without incurring antitrust liability, though such legislation should not address when or whether licensing is required in the first instance.

The framework further recommends that Congress consider establishing federal protections against the unauthorized distribution or commercial use of AI-generated digital replicas of an individual's voice, likeness, or other identifiable attributes—in order to protect individual privacy and liberty interests— while carving out clear exceptions for parody, satire, news reporting, and other First Amendment-protected expressive works. Finally, the framework urges Congress to continue monitoring the development of copyright precedents and enforcement in the courts and to evaluate whether additional legislative action may be necessary to address novel AI-related gaps or to provide additional protections for content creators depending on the outcomes of these decisions.

Read more here.

March 2026: Federal Deadlines That Will Reshape the AI Regulatory Landscape
Joe Cahill
On December 11, 2025, President Trump signed an Executive Order titled "Ensuring a National Policy Framework for Artificial Intelligence," declaring it the policy of the United States to achieve “global AI dominance through a minimally burdensome national policy framework for AI.”1 The Executive Order does not itself preempt, repeal, or invalidate any state AI law. Instead, it directs a sequenced series of federal agency actions that could result in litigation challenging state AI laws, new agency guidance on the application of existing federal statutes, and conditions on federal funding. Attorney General Pam Bondi established the Department of Justice's AI Litigation Task Force on January 9, 2026, tasked with challenging state AI laws in federal court on grounds including unconstitutional burdens on interstate commerce, preemption by existing federal regulations, and any other basis the Attorney General deems appropriate.2 But the Task Force has not yet filed any lawsuits — in part because the Executive Order contemplates a sequenced process in which the Department of Commerce first identifies specific state laws for referral. 

Read more here.

GSA's New AI Clause: Major Changes for AI Procurement
Parker Hancock
The federal government’s proposed AI procurement clause could force every company on a GSA Schedule to choose between its commercial AI terms and its government contracts. On March 6, 2026, GSA published “Basic Safeguarding of Artificial Intelligence Systems,” a clause that would rewrite the terms under which AI is sold to, or used in performance of work for, the federal government. The clause overrides commercial terms of service, claims government ownership of custom developments, restricts AI sourcing to American-made systems, and prohibits vendors from maintaining safety restrictions.

The reach extends well beyond AI-native companies. Any contractor that uses an AI-powered analytics platform, a machine learning-based cybersecurity tool, or even a coding assistant during contract performance could be affected. The terms are take-it-or-leave-it: MAS holders get 60 days to accept or lose schedule access. And OMB has declared compliance “material to contract eligibility and payment” — the phrase that could trigger False Claims Act liability, with treble damages and per-claim penalties for every non-compliant invoice.

Read more here.

U.S. Launches Call for Proposals for American AI Exports Program
Jason Wilcox
The Bottom Line: The U.S. Department of Commerce has announced the implementation of the American AI Exports Program, initiating a Call for Proposals from U.S. industry-led consortia to export full-stack AI technology packages. The 90-day submission window opens on April 1, 2026. Approved consortia will gain significant advantages, including expedited export license reviews and prioritized access to federal credit.

Background and Strategic Context
Driven by President Trump’s July 2025 AI Action Plan, the U.S. Department of Commerce’s International Trade Administration is rolling out a promotional program to secure America’s global leadership in artificial intelligence.

Following a public comment request published last October, the newly announced Call for Proposals targets “full-stack AI technology packages,” including AI-optimized computer hardware, data center storage, models, cybersecurity measures, and industry-specific applications.

Read more here.

Federal Court Signals Broader ADEA Exposure for AI‑Assisted Hiring
Leslie McCombs Roussev
A federal judge just cleared a major hurdle for the plaintiffs in Mobley v. Workday, allowing a high-profile AI hiring-bias case to move forward on core age-discrimination claims. In a March 6, 2026 order, Judge Rita Lin (N.D. Cal.) rejected Workday’s argument that the Age Discrimination in Employment Act (ADEA) only protects employees, not job seekers. This decision keeps the case’s central age-discrimination allegations in play as the litigation advances.

Judge Lin’s ruling was a split decision. The court dismissed several state-law claims and an individual plaintiff’s discrimination complaint, granting plaintiffs leave to amend by March 27, 2026 to address the identified deficiencies.

Zooming out, Mobley is part of a broader wave of legal and regulatory scrutiny of AI in hiring. As one example, a January 2026 proposed class action against Eightfold AI Inc. alleges that AI-generated hiring rankings should come with the same transparency required of credit bureaus. The Mobley and Eightfold lawsuits, together with emerging state AI employment laws, signal that legal frameworks are beginning to catch up with technological developments. Employers using AI-assisted screening should be prepared to explain what the hiring system does, how it is configured, and what monitoring exists to detect and remediate disparate impact.

Evolving Guidance on AI in EDVA Filings
Cailyn Reilly Knapp, Bradley Bennett
With the rapid adoption of generative artificial intelligence (“AI”) tools to research, edit, and draft court filings, courts have faced a new challenge: fabricated or inaccurate citations. In response, many federal judges, particularly in some of the country’s busiest commercial jurisdictions, have adopted standing orders addressing the use of AI in preparing filings. These orders aim to prevent parties from citing nonexistent authorities or mischaracterizing existing precedent.

These standing orders take different approaches. Some require parties only to verify the accuracy of AI-assisted work but do not mandate separate or additional filings.2 Others mandate that any party using an AI tool to generate any portion of a filing submit a separate declaration disclosing that use.

By contrast, judges in the EDVA, one of the country’s busiest and fastest-moving courts with a substantial commercial docket, have not adopted district-wide or judge-specific standing orders governing AI. The Court instead has addressed AI-related concerns through case-specific orders. Two recent examples come from the Alexandria and Richmond Divisions in cases involving pro se litigants. Although these orders stop short of establishing broad rules applicable to all litigants, they nevertheless reflect the Court’s longstanding expectations of candor, professionalism, and preparation.5 The technology may be new; the standards are not.

Read more here.

Quick Links
For additional insights on AI, check out Baker Botts’ thought leadership in this area:

1. How Data Centers Can Prep For Legal Challenges Amid War: Partners Derek Jones and Jonathan Gordon and Special Counsel Eric Robinson explain how recent military conflict in the Middle East has exposed data centers to state level physical threats, pushing them into the category of critical national infrastructure.
2. EU Committee on Legal Affairs Issues Report on Generative AI and Copyright: Senior Associate Nick Palmieri gives an overview of the report and next steps.
3. Oregon SB 1546: The First Chatbot Law With Real Teeth: Senior Associate Parker Hancock breaks down the new legislation, with practical next steps for companies impacted..
4. Supreme Court Denies Petition on Copyright Authorship by AI: Senior Associate Nick Palmieri provides a quick rundown on the decision.
5. AI Counsel Code Podcast
   - Legal Risks When Models Act on Their Own: Maggie Welsh speaks with Coleman Strine about the legal risks emerging as AI agents and AI‑powered browsers gain autonomy.
   - Houston as an AI Superpower: Inside the Rice Nexus Approach to AI Startups: Dr. Sanjoy Paul joins Maggie Welsh to break down what it really takes to move AI from research to real‑world deployment.