Pittsburgh-area Rep. Summer Lee is taking another swing at regulating AI discrimination.
Lee, along with several of her Democratic colleagues in the US House and Senate, announced plans on Tuesday to co-lead the reintroduction of the Artificial Intelligence Civil Rights Act — legislation that aims to hold tech companies accountable if they use discriminatory AI tools to make life-altering decisions for Americans.
The AI Civil Rights Act would ban algorithmic discrimination, or when an algorithm’s decision is unfair to a given user based on a class of protected traits.
“Discrimination is illegal when it’s done by a person and it should be just as illegal when it’s done by an algorithm,” Lee told Technical.ly. “These systems are already shaping life-changing decisions about jobs, housing, healthcare, loans and law enforcement, and too often they replicate the same racial, gender and disability-based inequities we’ve fought to dismantle for generations. Innovation cannot be allowed to outrun our obligation to protect civil rights and human dignity.”
The AI Civil Rights Act would ban algorithmic discrimination, or when an algorithm’s decision is unfair to a given user based on a class of protected traits, like race, gender or disability status. It would also require algorithm developers to “take reasonable measures” to prevent algorithm-induced harm.
That means companies would be required to conduct independent audits of their software. Plus, the bill would give Americans the right to choose whether an algorithm or a human is the final decisionmaker.
Under the legislation, the Federal Trade Commission would be responsible for enforcing this oversight.
While Congress remains divided, Lee said there’s a growing bipartisan recognition that AI needs real guardrails. In Western Pennsylvania, specifically, she said constituents have shared concerns about fairness, accountability and transparency when it comes to powerful technologies, which she intends to address.
“If technology is shaping people’s lives, then we have to mitigate the harms that it can cause, or in some cases are already causing,” Lee said. “We can’t allow companies to hide behind code when their products discriminate. People deserve transparency, real oversight and a human being who is beholden to civil rights law on the other side of life-changing decisions.”
Past efforts to regulate AI civil rights
A previous iteration of the Artificial Intelligence Civil Rights Act was introduced by Sen. Ed Markey in 2024 but ultimately died in committee. Markey now plans to reintroduce the companion bill in the Senate.
During the recent news conference, Markey cited a 2019 report that found mortgage lenders were 80% more likely to reject Black applicants, partially because of biased mortgage-approval algorithms. Another 2022 study found that AI models for predicting liver disease were twice as likely to miss the disease in women as in men.
Research has also found AI tools discriminate against women and people of color in hiring decisions.
“Bias is bias, whether it comes from a human or from a machine,” Markey said. “Under this bill, algorithmic discrimination will be unlawful, period.”
Lee’s record of fighting tech discrimination
This isn’t the first time Lee has tried to regulate AI bias.
Last year, she introduced the Eliminating Bias in Algorithmic Systems Act, legislation that would have established an Office of Civil Rights in each federal agency and monitored the agency’s systems for bias, discrimination and other harms caused by algorithms.
“AI bias is a real and pressing issue that’s impacting people right here in Pittsburgh,” Lee told Technical.ly at the time.
She cited local police deploying facial recognition tech to monitor Black Lives Matter protestors back in 2020 and the use of flawed algorithms in family welfare assessments.
The act died in committee last year.
The state of PA AI law
While there’s currently no federal legislation broadly regulating the development or use of AI, there are efforts to regulate the tech on a more local level in PA.
The commonwealth passed a law this summer making the use of AI for non-consensual deepfakes or voice clones a third-degree felony, and now the state assembly is working to require mandated reporters to also report AI-generated child sexual abuse material. On an even more local level, there have also been recent discussions on regulating AI in Philadelphia.
But there’s also federal efforts to stop this kind of AI regulation.
Earlier this year, a proposal to bar states from regulating AI for a decade ultimately failed to pass. Now, President Donald Trump is preparing to sign an executive order that would prevent AI regulation on the state level. However, the White House is likely to delay the order as Congress members consider adding similar legislation to the must-pass National Defense Authorization Act.
The order would not take immediate effect and would instead require further action by federal agencies to implement, likely triggering political and legal challenges by impacted states, according to experts.
