When the Hiring Algorithm Says No, Who Owns the Bias?

You fill out the form. Upload the résumé. Answer the questions. Then the rejection lands before a human being could finish their coffee.

That is the machinery behind a closely watched lawsuit against Workday, the company whose hiring software sits between millions of applicants and the jobs they want. A group of job seekers asked a federal judge last week to let their discrimination case move forward as a class action.

The lead plaintiff, Derek Mobley, is a Black man over 40 who has disabilities. He says he was rejected from more than 100 jobs at companies that used Workday’s hiring tools. The plaintiffs say the software disproportionately screened out African American, female, older and disabled applicants.

Those claims have not been proven. Workday denies wrongdoing and says its tools evaluate job qualifications, not protected traits like race, age or disability.

The machine did not have to “go rogue”

The lawsuit’s sharpest point is not that artificial intelligence suddenly became evil. The plaintiffs argue something more ordinary: software can work exactly as designed and still reproduce old discrimination if it learns from biased workplace patterns.

In 2024, U.S. District Judge Rita Lin allowed the case to proceed on a “disparate impact” theory. That legal doctrine covers policies that appear neutral but hit protected groups harder without enough business justification. The judge rejected claims that Workday intentionally designed the tools to discriminate.

That difference matters. Bias does not need a villain twirling a mustache. Sometimes it lives in the data, the screening rules and the decision to trust a score nobody applying for the job can see.

Who owns the rejection?

Workday says its software is customizable and employers decide how applicants are sorted, reviewed and rejected. The plaintiffs say Workday’s technology contributes to the outcome. That leaves a very modern accountability game: the employer points at the vendor, the vendor points at the employer, and the applicant is still unemployed.

The proposed case would include four subclasses: African Americans, women, people over 40 and people with disabilities who allegedly went through the challenged screening process. The judge has scheduled a class-certification hearing for March 9, 2027. Certification would not prove discrimination. It would decide whether the claims can be litigated together.

The scale is why this case matters. Reuters reports that Workday says two-thirds of Fortune 500 companies are customers. The plaintiffs point to company data showing more than 356 million applications moved through Workday Recruiting in 2024 alone. One case could shape how courts treat software vendors whose tools help decide who gets through the hiring gate.

Old law, new gatekeeper

The laws at issue are not new. The EEOC’s amicus brief identifies Title VII, the Age Discrimination in Employment Act and the Americans with Disabilities Act. What is new is the gatekeeper.

Black applicants have spent generations fighting hiring systems built around names, schools, ZIP codes, networks and “culture fit.” Putting those signals inside software does not make the question disappear. It can make the decision harder to see and easier to deny.

The latest class-action request and Workday’s response were reported by Reuters. The court has not decided who is right. But the case is already asking the question every automated system tries to dodge: when the machine says no, who has to answer for it?