AI Federal Framework

State of play.

• DOJ has intervened as a formal party in the Colorado algorithmic discrimination case, obtaining a stay of SB24-205 enforcement on Equal Protection grounds — the first time the federal government has directly litigated to block a state AI anti-discrimination law (→ DOJ Intervenes in xAI Lawsuit to Block Colorado's AI Discrimination Law[1][2][3]).
• The White House is moving toward mandatory pre-release AI review, with every major US frontier lab now participating in voluntary CAISI assessments and an executive order mandating industry-wide review under active consideration (→ Google, Microsoft and xAI Agree to Share Early AI Models With U.S., White House pushes federal AI review standards to eliminate "ideological bias").
• Federal procurement standards have hardened through the "Preventing Woke AI" EO, which requires agencies to implement "Unbiased AI Principles" for LLM procurement, with OMB implementing guidance and contract revision obligations on a rolling 90-day clock (→ White House pushes federal AI review standards to eliminate "ideological bias").
• The National AI Legislative Framework's preemption architecture is now being stress-tested in litigation, with the Colorado case as the template — DOJ's task force is positioned to replicate the intervention against other state AI laws (→ DOJ Intervenes in xAI Lawsuit to Block Colorado's AI Discrimination Law[1][2][3], Trump Admin Releases National AI Framework on March 20, 2026).
• For counsel advising AI developers, federal contractors, or enterprise deployers, the practical baseline has shifted from "watch and wait" to active compliance planning: preemption is being litigated now, federal procurement standards are attaching, and voluntary pre-release review is hardening toward mandatory.

Where things stand.

• The March 20, 2026 National Policy Framework is the structural anchor. The White House legislative recommendations propose federal preemption of state AI laws imposing "undue burdens," sector-specific oversight through existing agencies (FDA, CMS, DOJ, FTC, Commerce), and no new dedicated AI regulator (→ White House Releases National AI Policy Framework on March 20, 2026, Trump Admin Releases National AI Framework on March 20, 2026, White House Releases 2026 National AI Policy Framework on March 20). The framework's full statutory language remains unpublished; no preemption legislation has passed.
• State law continues to attach. Over 600 AI bills were introduced in state legislatures in Q1 2026 alone; Colorado's SB24-205 (now stayed), California's AB 3030, Utah's AI Act, and Illinois psychotherapy restrictions remain on the books pending federal action (→ Alston & Bird Publishes April 2026 AI Quarterly Review of Key U.S. Laws and Policies). The Commerce Department evaluation of "onerous" state laws was due March 11 but remains unreleased.
• DOJ's preemption task force is operational. The Colorado intervention is the first exercise of the task force authority established in the December 2025 EO and the March Framework — with Commerce evaluating state laws and DOJ litigating against them (→ DOJ Intervenes in xAI Lawsuit to Block Colorado's AI Discrimination Law[1][2][3]).
• Pre-release government model review is consolidating. CAISI (under NIST/Commerce) has completed more than 40 assessments; Google, Microsoft, and xAI have now joined OpenAI and Anthropic in the voluntary program, with safety safeguards reduced for government testing (→ Google, Microsoft and xAI Agree to Share Early AI Models With U.S.).
• Federal procurement AI standards are live. The "Preventing Woke AI" EO requires "Unbiased AI Principles" in agency LLM procurement; OMB guidance and contract revision obligations are on a 90-day implementation clock (→ White House pushes federal AI review standards to eliminate "ideological bias").
• The Anthropic-White House standoff has structurally thawed. Anthropic's CEO met with the White House Chief of Staff and Treasury Secretary; OMB is developing a safeguards framework for federal agency access to Mythos (→ Anthropic CEO Amodei Meets Trump Officials on Mythos AI Risks[1][3]). The underlying leverage — supply-chain-risk designation and Defense Production Act threats — has not been formally withdrawn.
• Sectoral regulators are building parallel frameworks. The USPTO has a five-pillar AI strategy, updated inventorship rules barring AI as independent inventor, and a generative AI tool deployed to over 200 examiners (→ BakerHostetler Podcast on USPTO's AI Strategy and Guidance Evolution[12][15]). Healthcare sector analysis identifies FDA, CMS, and DOJ as the operative regulators under the Framework (→ White House Releases 2026 National AI Policy Framework on March 20).
• The Trump America AI Act discussion draft would codify the December 2025 EO's preemption directive; H.R. 8094 (AI Foundation Model Transparency Act) imposes disclosure-only obligations on LLM developers — the two legislative vehicles are not yet reconciled on copyright liability or Section 230 (→ Alston & Bird Publishes April 2026 AI Quarterly Review of Key U.S. Laws and Policies).
• The White House cybersecurity vulnerability-patching initiative — led by National Cyber Director Cairncross — is the operational complement to CAISI assessments, focused on identifying and patching AI model vulnerabilities before public release .
• IP ownership for AI-generated code is an emerging litigation front. Autonomous coding agents generating production-ready software have triggered debate over whether ownership vests in the user, the developer, or neither — the National AI Framework has begun addressing training-data IP and fair use but without settled statutory language (→ Q1 2026 AI Agents Spark IP Debates in Software Development).

Latest developments.

• DOJ intervenes in Colorado SB24-205 litigation, obtains enforcement stay; Colorado task force faces May 13 deadline for successor legislation (→ DOJ Intervenes in xAI Lawsuit to Block Colorado's AI Discrimination Law[1][2][3])
• Google, Microsoft, and xAI join CAISI pre-release review program; every major US frontier lab now participating; mandatory review EO under consideration (→ Google, Microsoft and xAI Agree to Share Early AI Models With U.S.)
• White House "Preventing Woke AI" EO establishes "Unbiased AI Principles" for federal LLM procurement; OMB guidance and contract revision obligations on 90-day clock (→ White House pushes federal AI review standards to eliminate "ideological bias")
• Autonomous AI coding agents generating production-ready software trigger IP ownership and developer liability debate; National AI Framework begins addressing training-data IP without settled language (→ Q1 2026 AI Agents Spark IP Debates in Software Development)
• National AI Framework healthcare analysis identifies FDA, CMS, and DOJ as operative sectoral regulators; 177 state AI bills pending across 31 states as of April 2026 (→ White House Releases 2026 National AI Policy Framework on March 20)
• UK FCA reaffirms principles-based AI oversight through existing Consumer Duty and SM&CR framework; Mills Review underway on autonomous system accountability (→ FCA Sticks to Existing Rules for AI Oversight in Finance)

Active questions and open splits.

• Whether the Colorado intervention becomes the federal template. DOJ's Equal Protection theory — that disparate-impact liability compels demographic adjustments — is novel and untested at the appellate level. If it holds, it forecloses most state algorithmic bias statutes; if it fails, state enforcement authority survives (→ DOJ Intervenes in xAI Lawsuit to Block Colorado's AI Discrimination Law[1][2][3]).
• Whether voluntary pre-release review hardens into mandatory obligation. Every major lab is now in the CAISI program voluntarily; an EO mandating industry-wide review is under active consideration. The line between voluntary and compelled — and what triggers liability for non-participation — is unresolved (→ Google, Microsoft and xAI Agree to Share Early AI Models With U.S.).
• Preemption scope: development vs. use. The Framework's proposed distinction — federal authority over AI development, state authority over AI use — has not been tested under the major questions doctrine. Whether courts accept that line, and how it maps onto state anti-discrimination and disclosure statutes, is the central constitutional question (→ Trump Admin Releases National AI Framework on March 20, 2026).
• Federal procurement "ideological neutrality" standard. The "Unbiased AI Principles" requirement creates a procurement compliance obligation whose content OMB has not yet defined. Vendors face the risk of contract revision requirements against an undefined standard (→ White House pushes federal AI review standards to eliminate "ideological bias").
• IP ownership for AI-generated outputs. No settled rule exists on whether code, analysis, or other outputs generated by autonomous AI agents vest in the user, the developer, or remain unowned. Litigation is anticipated; the National AI Framework addresses the question without resolving it (→ Q1 2026 AI Agents Spark IP Debates in Software Development).
• Open-source vs. proprietary in national security context. The USCC report showing Chinese open-source models leading on HuggingFace has sharpened the policy debate; the administration has not resolved whether to embrace open models or treat them as a security risk — and the answer directly affects export controls and procurement .
• EU AI Act August 2026 binding effect vs. US framework divergence. Multinational clients face a compliance fork: the EU's risk-based prescriptive regime binds in August while the US framework remains non-binding legislative recommendations. Whether the two converge on specific standards — or require parallel compliance architectures — is unresolved (→ Alston & Bird Publishes April 2026 AI Quarterly Review of Key U.S. Laws and Policies).

What to watch.

• Colorado's May 13 deadline for successor legislation — whether the amended statute addresses DOJ's Equal Protection theory or triggers a second intervention.
• Whether the White House issues an EO mandating industry-wide pre-release AI review, and what enforcement mechanism it carries.
• OMB publication of the "Unbiased AI Principles" implementing guidance and what it requires of existing federal AI contracts.
• Whether DOJ's preemption task force files additional interventions against California, New York, or other state AI statutes.
• EU AI Act binding effect in August 2026 — whether US developers treat it as a compliance floor that implicitly shapes domestic practice.
• Commerce Department release of its evaluation of "onerous" state AI regulations — the document that was due March 11 and remains unreleased.