§ A.I. · 3 MIN READ
A Robot Read Your Resume. A Judge Noticed.
In June a federal judge ruled that the AI screening software used by most of the Fortune 500 must face class-action claims that it discriminated. Here is what that means for the next job you apply to.
By Culture
THE SHORT VERSION
In June 2026 a federal judge in California ruled that Workday, whose AI-powered screening software is used across the Fortune 500, must face class-action claims alleging its algorithms discriminated against Black applicants, women and older workers. A separate Stanford-led study of about 4 million applications found clear racial disparities, affecting more than a quarter of Black applicants.
For years the advice was to make your resume stand out. That advice assumed a person was reading it. For most jobs at most large companies, the first reader is now software, and this summer a court said out loud what applicants have suspected for a decade.
In June 2026 a federal judge in California ruled that Workday, whose AI-powered screening software is used across virtually the entire Fortune 500, must face class-action claims alleging its algorithms discriminated against Black applicants, women and older workers. That is not a verdict. It is a ruling that the case is serious enough to proceed as a class, which is the door most discrimination claims never get through.
The research arrived alongside it. A Stanford-led study of roughly four million job applications found clear racial disparities in AI hiring tools, affecting more than a quarter of Black applicants.
A screening model learns from who a company hired before. If the people it hired before were mostly one kind of person, the model learns that kind of person is what success looks like, and it filters accordingly. Nobody has to type a racist rule for the outcome to be a racist rule. That is what makes it hard to see and hard to sue over.
It is the same mechanism in lending. Urban Institute analysis of federal mortgage data found Black and Brown borrowers were more than twice as likely to be denied a loan than white borrowers, and in 2022 Wells Fargo was accused of using a creditworthiness algorithm that scored Black and Latino applicants as riskier than white applicants with similar finances.
“Nobody has to write a racist rule. They only have to train on a racist past and call the output objective.”
Federal action has largely stalled, so states are writing the law. Colorado's AI Act is the first comprehensive state statute on algorithmic bias, in force from June 30, 2026. Expect your own state to follow, and expect the fight to be about whether anybody outside a company is allowed to audit its model.
That is worth paying attention to, because the same shift is deciding who gets funded and what a credit score is allowed to weigh. If a machine is going to judge you, somebody should be allowed to check its work.
- Did a court rule that AI hiring software is discriminatory?
- Not yet. In June 2026 a federal judge in California ruled that Workday must face class-action claims alleging its AI screening discriminated against Black applicants, women and older workers. That allows the case to proceed as a class; it is not a final finding of liability.
- How do I get my resume past an AI screener?
- Mirror the exact words in the job posting, since screeners match text rather than meaning. Send a single-column document with standard headings and no tables, graphics or text in the header. Spell out acronyms both ways. Apply on the company site as well as the job board, and find one human for a referral, because referrals usually enter a different queue.
- Why is AI hiring software biased?
- Because it learns from who a company hired in the past. If past hires skewed toward one group, the model treats that profile as what success looks like and filters for it. No one needs to write a discriminatory rule for the result to be discriminatory.
- Is there a law against AI bias?
- Federal action has largely stalled, so states are acting. Colorado's AI Act is the first comprehensive state law on algorithmic bias and took effect June 30, 2026. Other states are following.
