The US Justice Department intervened on April 24, 2026, in xAI’s federal lawsuit challenging Colorado’s AI antidiscrimination law, making it the first time the Trump administration has moved to block a state AI regulation through the courts.

The move adds significant legal weight to what was already a high-profile case. xAI, Elon Musk’s AI company, filed its lawsuit on April 9 seeking an injunction to prevent Colorado from enforcing Senate Bill 24-205 before its June 30, 2026 effective date. The DOJ filed a Statement of Interest in support of xAI’s challenge, arguing that the Colorado law itself is unconstitutional.

What the DOJ Is Arguing

The Justice Department’s core legal argument targets Colorado’s use of disparate-impact liability. The Colorado AI Act requires developers and deployers of “high-risk” AI systems to take reasonable care to prevent algorithmic discrimination based on protected characteristics like race and sex. The DOJ argues this mechanism violates the Equal Protection Clause of the Fourteenth Amendment.

Specifically, the government alleges that “by imposing disparate-impact liability based on statistics alone,” the law forces AI companies to engage in “demographic-conscious engineering” of their models, calibrating outputs to eliminate statistical disparities rather than intentional discrimination. The DOJ further takes issue with the law’s exemption for certain forms of discrimination designed to advance “diversity,” calling that carve-out constitutionally impermissible.

This line of argument tracks closely with the Trump administration’s broader stance on diversity-related requirements, and it signals that the White House views state-level AI antidiscrimination mandates as falling under the same constitutional scrutiny it has applied to corporate DEI programs.

The Broader Policy Context

The DOJ’s intervention is not just about Colorado. According to reports citing the filing, the government argues this case falls under President Trump’s Executive Order 14365, which directed federal agencies to discourage, challenge, and potentially preempt state AI laws that conflict with the administration’s approach.

This makes Colorado a test case. If the federal courts agree that states cannot impose AI liability standards that require demographic-conscious model adjustments, it would create a template for challenging similar AI regulations in other states. Texas, Illinois, Tennessee, Washington, and several other states have enacted or are debating their own AI governance laws.

Separately, the Colorado Attorney General’s office quietly delayed enforcement of the law, and the state’s AI Policy Work Group released a proposal in March to replace much of the original statute with a streamlined framework. If enacted, that proposal would also push the effective date back to January 1, 2027. That work is ongoing, and no final legislation has been passed.

For now, SB 24-205 remains on track to take effect June 30, 2026 unless a court grants an injunction.

What This Means for Business

If your company deploys AI systems in Colorado or other states with active AI regulations, this situation deserves close attention.

In the near term, the legal uncertainty around Colorado’s law gives businesses reason to monitor the court proceedings before investing heavily in compliance infrastructure. If a federal court grants an injunction or the legislature rewrites the law, the compliance picture shifts significantly.

In the longer term, the DOJ’s intervention signals something important: the federal government is not planning to leave state AI regulation alone. The trajectory points toward either federal preemption, where Congress passes a national AI framework that overrides state laws, or a prolonged period of patchwork state regulation under federal legal challenge. Neither scenario is particularly stable for businesses trying to plan AI deployments.

The practical advice for most companies is still the same: document your AI use cases, understand which ones qualify as high-risk under any of the emerging definitions, and build out risk management and disclosure frameworks. That work is useful under any regulatory regime, and it protects you in the interim regardless of how the legal fights resolve.

What’s changing is the scale of the uncertainty. A few months ago, Colorado’s law looked like a compliance deadline. Now it’s a live constitutional fight with the federal government as a participant.

Companies using AI for employment decisions, credit screening, healthcare, housing, or education are the ones most directly exposed. If you fall into those categories, now is the time to talk to counsel, not wait for the dust to settle.