Workday has not lost the AI hiring case, but the lawsuit has already done something important: it has made the software vendor part of the discrimination fight, not just the employers that use its tools.
If you sell hiring software that screens people before a manager ever opens a resume, Mobley v. Workday is the case you should be reading. U.S. District Judge Rita Lin allowed Derek Mobley's federal age discrimination claim to proceed as a collective action in May 2025, and that ruling has turned a frustrated job search into a real test of who answers when an algorithm filters out applicants at scale.
Mobley, a Black IT worker over 40 who has anxiety and depression, sued Workday in 2023 after applying for more than 100 jobs through employers using Workday's recruiting platform and getting nowhere. The Wall Street Journal reported that some rejection emails arrived in the middle of the night or within an hour, which is exactly the kind of detail that makes job seekers suspect a machine made the call before a human did.
Workday denies the claims. The company has said its software matches resume keywords to qualifications set by employer customers, scores candidates as strong, good, fair or low matches, and leaves final hiring decisions to those employers. That defense is straightforward enough: we provide the tool, the customer decides. Judge Lin has already refused to let that framing end the case.
That is the point every AI hiring vendor needs to understand. If your product ranks, scores, recommends, screens, or helps reject candidates, you are not selling office stationery. You are sitting inside the hiring process. Courts and regulators are now asking whether that position carries civil rights obligations of its own.
Bloomberg Law reported in May 2025 that the Workday bias suit could go forward as an age-claim collective action under the Age Discrimination in Employment Act. Later reporting noted that Workday was ordered to provide information about clients using AI features in hiring. Those are procedural steps, not a final verdict, but they matter because they force more of the system into view. The black box is less useful as a business model once a judge starts asking who used it and how.
The case is still narrower than some commentary makes it sound. The court has not found that Workday discriminated against Mobley. According to the Journal, Judge Lin had already ruled that Workday did not intentionally discriminate against him, while leaving room for him to try to prove that the technology had a discriminatory effect. That distinction matters. Intent is one thing. Disparate impact is another. A vendor can lose the second fight even while insisting nobody inside the company set out to discriminate.
For enterprise buyers, the practical problem is ugly and immediate. If you use software from Workday, HireVue, Eightfold, Pymetrics, or any similar HR-tech vendor, you cannot point at the contract and assume the risk lives somewhere else. The EEOC has made clear that employers can be liable when vendor-built AI tools produce discriminatory outcomes under federal employment law. You bought the system. You used the output. That is enough to bring you into the argument.
California has made the pressure sharper. The Washington Post reported in December 2025 that California rules now confirm anti-discrimination protections apply to automated decision systems, including AI, machine learning, and algorithmic tools used in employment decisions. If you are screening candidates with automation and keeping weak records, you are creating the evidence problem before the lawsuit even arrives.
Look at the contracts too. A standard SaaS liability cap worth twelve months of fees is not much comfort if thousands of rejected applicants claim a tool screened them out unlawfully. Procurement teams should be asking for audit rights, bias testing documentation, clear indemnity language, and explanations of what data the system uses to score people. Do not wait until discovery to learn how your hiring funnel works.
None of this means automated hiring tools are going away. Large employers use them because applicant volume is enormous, and manual screening for every role is not coming back. But vendors have had a long run selling speed while treating bias testing as a trust-and-safety feature somewhere off to the side. That posture is finished.
The Workday case is current because it sits at the center of a live legal question: when software helps decide who gets seen, who gets rejected, and who never hears back, who is responsible for the outcome? The answer will not come from a product brochure. It will come from court orders, regulator guidance, and the records companies kept when they thought the tool was doing routine work.
Also read: Baseten raises $1.5 billion at a $13 billion valuation as inference becomes AI's most contested infrastructure layer • Insilico Medicine signs a $2.5 billion AI drug discovery deal with SK Biopharmaceuticals for neuroimmune disorders • Oracle cut 21,000 jobs to fund its $50 billion AI buildout and workers say they trained their own replacements
