For generations, civil rights battles centered on people: employers who refused to hire qualified Black applicants; banks denying mortgages in minority neighborhoods; landlords steering families away from certain communities.
Now, civil rights advocates say the next fight may be against algorithms.
As artificial intelligence (AI) becomes increasingly involved in decisions affecting jobs, housing, lending, health care and public benefits, a coalition of consumer advocates and civil rights organizations says the Trump administration is making it harder for states to stop technology from repeating decades of discrimination.
“Testing and evaluation of AI systems, including removing bias, make systems more accurate, not less,” said Odette Williamson, director of racial justice advocacy at the National Consumer Law Center. “The FTC’s proposed policy statement creates confusion about how to legally assess bias and discrimination in AI, and it undermines states’ ability to protect consumers from discriminatory AI systems. It should be withdrawn in its entirety.”
The warning comes as the administration pursues two separate efforts that critics say could weaken oversight of artificial intelligence.
Last week, the National Consumer Law Center filed comments urging the Federal Trade Commission to withdraw a proposed policy statement that the organization says discourages companies from testing AI systems for racial and other forms of bias. Days later, President Donald Trump signed an executive order directing federal agencies to examine ways to prevent states from enforcing or adopting AI laws that the administration believes interfere with a national approach to artificial intelligence.
Together, those actions have intensified a debate that extends well beyond Silicon Valley. The dispute now reaches into questions that have defined civil rights policy for decades: Who gets approved for a mortgage? Who is hired for a job? Who qualifies for housing? Who receives government benefits? Increasingly, advocates say, software—not people—is helping make those decisions.
Artificial intelligence already plays a role in screening job applicants, evaluating loan applications, recommending tenants, detecting fraud, assisting insurers and supporting health care decisions. Consumer advocates argue that when those systems are trained on historical data that contain racial or economic disparities, they can reproduce those outcomes unless developers continually test for bias and correct it.
“Opaque AI systems need guardrails and oversight to make sure they do not discriminate against people based on race, age, gender, religion, and many other factors,” Williamson said.
The National Consumer Law Center argues that the FTC proposal rests on a false premise—that efforts to reduce bias somehow make artificial intelligence less accurate. Instead, the organization says continuous testing is necessary to improve accuracy and comply with existing anti-discrimination laws. It also argues that the commission has failed to show evidence that companies attempting to comply with state civil rights protections are deceiving consumers.
“With Congress unwilling to pass legislation protecting people from AI and this administration gutting our consumer protection agencies, states are essential to stop AI from unleashing digital redlining,” Williams continued.
‘States Have a Vital Role in Protecting the Public from … Unknown Risks of AI’
Civil rights organizations have expanded that argument beyond the FTC.
In a letter sent to leaders of the House Democratic Commission on AI and the Innovation Economy, the Leadership Conference on Civil and Human Rights and dozens of national organizations urged lawmakers to reject any proposal that would override state AI protections. The groups called for mandatory bias testing, privacy protections, enforcement authority for federal and state regulators, and preservation of disparate-impact protections under civil rights law.
The coalition argued that public concern about AI regulation spans political affiliations. It cited polling showing broad support for requiring companies to demonstrate that AI systems used in employment, lending and housing do not discriminate. The organizations also warned that communities of color could face disproportionate economic consequences as AI reshapes the workforce and automates more occupations.
The administration has argued that a patchwork of state AI laws could complicate innovation and create inconsistent regulatory standards nationwide. The executive order specifically identifies California’s AI disclosure requirements and Colorado’s law addressing algorithmic discrimination as examples of state measures federal agencies should examine.
Technology companies have likewise argued that complying with dozens of different state AI laws could become increasingly difficult as artificial intelligence expands across industries.
Advocates counter that states have historically acted first when Congress has been slow to respond to emerging risks.
“States have a vital role in protecting the public from the myriad of known and unknown risks of AI, which can be used to improperly reject people for credit, jobs, and housing; freeze or steal bank accounts; abuse and share private data; and raise the cost of living through surveillance pricing,” said Lauren Saunders, associate director and director of federal advocacy at the National Consumer Law Center.
She said efforts to prevent states from regulating artificial intelligence reverse a long tradition of state consumer protection.
“States on a bipartisan basis have been addressing the risks of AI, exercising their role as our first responders and laboratories of democracy, which is critical given Congress’s inability to address the dangers of AI,” Saunders said. “Our Constitution limits the powers of the federal government and preserves the power of the states. The actions that President Trump is directing federal agencies to take to block states from protecting people from current and future AI dangers are reckless, unlawful, and will not stand up to challenge.”
