Musk, DOJ Sue Colorado Over AI Bias Law; Arguments Flawed, Expert Says

The U.S. Department of Justice joined Elon Musk’s artificial intelligence company xAI in an April lawsuit against Colorado, seeking to invalidate the state’s AI anti-discrimination law.

When the federal government aligns with a billionaire against a state attempting to protect residents from AI discrimination, the implications extend beyond Colorado, affecting the entire nation.

The Justice Department’s lawsuit is part of a coordinated federal effort to characterize AI consumer protections as ideological overreach. In July 2025, President Donald Trump signed an executive order on “preventing woke AI,” equating bias mitigation measures to a leftist agenda that suppresses free speech and truth. The federal National Policy Framework, launched in March, included a push to preempt state AI laws, specifically targeting Colorado’s law. The Justice Department’s intervention marks the first time the federal government has sought to intervene in a lawsuit challenging a state AI statute.

The Colorado bill in question, Senate Bill 205, aimed to protect individuals from discrimination in high-risk AI systems that make consequential decisions regarding hiring, housing, and healthcare.

The original legislation required bias audits, impact assessments, and disclosure, which drew opposition from the business community. Colorado responded by revising the requirements in mid-March, adopting a reduced transparency framework before xAI filed its lawsuit.

Such legislation matters. A 2019 study published in Science revealed that a widely used healthcare algorithm deployed across U.S. hospitals assigned Black patients half the care of equally sick white patients, using healthcare costs as a proxy for health needs and embedding existing inequities into clinical decisions. Removing healthcare costs eliminated racial bias in the model and improved care delivery.

xAI argued that the Colorado bill would force the company to promote the state’s “ideological views on various matters, racial justice in particular,” in its Grok chatbot. State Representative Brianna Titone, a lead sponsor of the bill, countered: “SB 205 is about consequential decisions. We’re not restricting speech. Our bill does not say that Grok still can’t be a dick.”

The Justice Department characterized the bill as “state-mandated discrimination” that “obligates AI developers and deployers to discriminate,” an extraordinary description of a law requiring companies to check whether their systems produce unlawfully discriminatory outcomes and correct them.

The department’s argument rested on the false assumption that AI systems use “neutral criteria.” Under that logic, altering models that use neutral criteria would compromise their objectivity, constituting discrimination. However, as the hospital study shows, criteria that appear neutral can produce worse outcomes for certain populations. This pattern of poor proxies harming marginalized groups has been documented in studies on welfare allocation and college admissions. Bias also arises when models have less training data on some groups, causing worse performance for those populations, as shown in facial recognition research and studies on large language models. When companies improved representation in their training data, performance discrepancies nearly disappeared.

The “winning the AI race” framing, reflected in the Justice Department lawsuit, conflates the pursuit of artificial general intelligence and beating China with rules barring AI discrimination in hiring and healthcare. These are separate issues, and conflating them is misleading.

The Wall Street Journal, the only mainstream national outlet to cover the story, focused on the law’s potential impact on businesses and “killing the entrepreneurial spirit,” a sentiment echoed in the Justice Department lawsuit. However, no companies have been cited as leaving Colorado because of the regulation. Governor Jared Polis stated that “far more” firms are moving to Colorado than leaving. The Journal’s only concrete example was Palantir, which cited the bill as a possible burden in an SEC filing but did not say it was the reason for moving to Florida. Conversely, companies including Microsoft have flagged AI bias and discrimination as a material risk.

Small businesses have legitimate concerns, and the state responded. Still, using AI to make consequential decisions about hiring or healthcare is a choice. If a business lacks the bandwidth to check whether its AI discriminates, it lacks the bandwidth to use it responsibly. AI tools that fail to work for everyone deliver worse outcomes, create liability, and erode trust. The healthcare algorithm that was fixed did not sacrifice efficiency; it improved performance.

The pressure campaign succeeded. On May 14, Polis signed into law a new bill, SB 189, which repeals and replaces SB 205. Gone are requirements for companies to proactively assess high-risk AI systems for bias, conduct annual reviews, report discovered discrimination to the state, and use reasonable care to address known and foreseeable harms. What remains is a requirement for AI developers to share technical documentation with deployers, but not the public. Consumers receive notice that AI was involved in a decision and the right to request human review, though most consumers will never know to invoke it. The new law is better than nothing but falls short of ensuring meaningful, proactive accountability for high-risk systems.

If the Justice Department can join a billionaire’s lawsuit to kill a state law meant to protect consumers and support transparency in high-risk areas, the message to other states is clear: do not try.

That approach is not pro-America. AI is an incredibly powerful technology, and its potential is greatest when it works for everyone. We deserve the transparency and protection to achieve that.

Dr. Genevieve Smith is a postdoctoral research fellow at Stanford University, founder of the Responsible AI Initiative at the UC Berkeley AI Research Lab, and a member of the professional faculty at the UC Berkeley Haas School of Business.