A Judge Freezes Colorado's Landmark AI Bias Law After a Challenge From Musk's xAI

Colorado spent two years building the country's most ambitious law against discriminatory artificial intelligence. It took Elon Musk's xAI barely two weeks to help freeze it.

On April 9, 2026, xAI filed a civil complaint in the U.S. District Court for the District of Colorado, x.AI LLC v. Weiser, seeking to permanently enjoin Senate Bill 24-205 — the Colorado AI Act — as unconstitutional. Fifteen days later the U.S. Department of Justice moved to intervene on the company's side, the first time the federal government has gone to court to invalidate a state AI statute. And on April 27, a federal magistrate judge granted a joint motion to stay enforcement of the law, halting a measure that was set to take effect June 30. As of July 2026, that freeze remains in place, and Colorado's landmark statute has effectively been legislated and litigated into suspended animation.

What the law required

Passed in May 2024 and billed as the first comprehensive state AI antidiscrimination law in the nation, SB 24-205 targeted "high-risk" AI systems — those that make, or are a substantial factor in making, "consequential decisions" in areas such as employment, housing, lending, insurance and education. It required developers and deployers of those systems to use reasonable care to protect consumers from "algorithmic discrimination," defined as unlawful differential treatment or impact based on protected characteristics like race, sex or age.

The obligations were concrete. Developers had to disclose how a system was trained and what discrimination risks its intended uses might create. Deployers had to adopt a risk-management program, conduct impact assessments, and notify consumers when an AI system played a role in a consequential decision affecting them. Governor Jared Polis signed the bill while openly warning it could burden the state's tech sector, and lawmakers twice pushed back its effective date — originally February 2026, then June 30 — as an AI Policy Work Group hunted for a compromise that never fully materialized.

The challenge and the stay

xAI, the Musk-owned developer of the Grok chatbot, attacked the statute on four fronts: that it violates the First Amendment by compelling speech and discriminating on content and viewpoint, burdens interstate commerce under the Commerce Clause, is unconstitutionally vague, and violates the Equal Protection Clause. In its complaint the company called the Act "an effort to embed the State's preferred views into the very fabric of AI systems," describing it as "controversial and legally suspect." Its provisions, xAI argued, "prohibit developers of AI systems from producing speech that the State of Colorado dislikes, while compelling them to conform their speech to a State-enforced orthodoxy on controversial topics of great public concern."

xAI reserved particular fire for the law's carve-out excluding outputs meant to expand diversity in an applicant or customer pool or to redress historical discrimination — a provision the DOJ later seized on. The federal government's complaint in intervention argued the Act violates equal protection two ways at once: that it "compels discrimination" through disparate-impact liability, "distort[ing] AI model outputs" so developers must account for race, sex and religion, and that it simultaneously "authorizes discrimination" by exempting diversity-promoting systems under Section 6-1-1701(1)(b).

Assistant Attorney General Harmeet K. Dhillon framed the intervention in stark ideological terms, accusing Colorado of trying to "coerce our nation's technological innovators into producing harmful products that advance a radical, far left worldview at odds with the Constitution." Assistant Attorney General Brett A. Shumate said the law threatened national and economic security by promoting "ideological bias." Colorado Attorney General Philip Weiser, the named defendant and himself a longtime skeptic who has called the law "problematic," declined to comment on the suit.

Rather than fight the timeline, the parties struck a deal. Magistrate Judge Cyrus Y. Chung's April 27 order — issued on a joint motion by xAI and state regulators — stays enforcement and provides that xAI will file a motion for preliminary injunction or an amended complaint within 28 days after Colorado adopts either implementing rules or new legislation amending or replacing the Act. Conduct occurring on or before the 14th day after the court rules on that injunction is shielded from penalties and investigation.

Why It Matters

This is the opening skirmish in a federal-versus-state war over who governs AI. After Congress failed to pass a moratorium on state AI laws, President Trump's December 2025 executive order singled out Colorado's Act by name and directed the DOJ to stand up an AI Litigation Task Force to attack state statutes on Commerce Clause, preemption and other grounds. The xAI intervention is the strategy in action: rather than wait for federal legislation, the administration is using the courts — and a friendly private plaintiff — to knock state laws down one at a time. A win in Colorado would hand a template to challengers eyeing AI rules in Illinois, Texas and California, where xAI is already fighting the AB 2013 training-data law before the Ninth Circuit.

What to Watch

Colorado did not wait for a verdict. On May 14, Polis signed SB 26-189, the Automated Decision-Making Technology Act, a narrower, disclosure-focused replacement that repeals and reenacts the old framework and pushes any effective date to January 1, 2027. Because the stay explicitly covers legislation "replacing or amending" SB 24-205, Weiser has said his office will not enforce the new law until rulemaking concludes — meaning the litigation clock resets around that process. Watch for when xAI files its promised preliminary-injunction motion, whether the court reaches the constitutional merits or lets the case dissolve into the rewritten statute, and whether Commerce ever publishes its long-overdue list of "onerous" state AI laws. Each is a signal of how aggressively Washington intends to preempt the states.