A federal judge in California has cleared the way for one of the most closely watched AI hiring discrimination cases to target the screening technology itself, not just the employers who deployed it, a shift that puts AI vendors squarely in the blast radius of employment litigation for the first time.

On June 22, 2026, Judge Rita F. Lin of the U.S. District Court for the Northern District of California ruled on Workday’s motion to dismiss in Mobley v. Workday (No. 23-cv-00770), the case brought by job applicants Derek Mobley, Jill Hughes, Sheilah Johnson-Rocha, and FaithLinh Rowe over the company’s AI-powered applicant screening tools. The order denied Workday’s bid to throw out the plaintiffs’ claims under California’s Fair Employment and Housing Act (FEHA), a decision that reaches beyond California’s borders.

What the court decided

The plaintiffs allege that Workday’s algorithmic screening products, including its AI recommendation engine, filtered out qualified candidates on the basis of race, age, and disability, at scale, across the hundreds of employers that license the platform. Workday argued the FEHA claims should be dismissed because the company is a technology vendor, not an employer, and that liability should attach only to the businesses making hiring decisions.

Advertisement

HRTech Your brand belongs here. Reach the decision-makers who read HRTech every day. Premium placements across the site and newsletter. Advertise with us

Judge Lin rejected that argument. The court found the plaintiffs adequately alleged that Workday’s screening tools are designed, developed, and operated from the company’s California headquarters, which the order said was enough to establish California’s jurisdictional nexus over discrimination claims, including for applicants who were never physically in the state and whose employer never was either. The court also allowed one plaintiff, Hughes, to proceed on a proxy-discrimination theory under the Americans with Disabilities Act, finding her claim that the algorithm used indicators correlated with disability fell within the scope of an earlier amendment.

Not every claim survived. Rowe’s race-based disparate-impact claim was dismissed because she had not sought the court’s permission to add a race-discrimination theory, and a broader argument that Workday should face direct-employer liability was struck for the same procedural reason. Those dismissals leave room for amendment, meaning the scope of the case could still widen.

Why the California nexus theory matters beyond one company

The significance of the ruling is not really about Workday. It is about the legal theory the court endorsed: that an AI vendor can be sued under state discrimination law based on where its software is built and run, regardless of where the harmed applicant or the hiring company is located. If that theory holds through trial, or survives appeal, it becomes a template for reaching any hiring-technology company headquartered or developing its models in California, which is to say most of the sector.

That is a meaningfully different exposure than the one HR and legal teams have been bracing for. As we reported when AI hiring and layoff tools first became employers’ biggest litigation exposure, most compliance planning has assumed the employer using the tool absorbs the legal risk, with vendor contracts and indemnification clauses treated as a private matter between the two parties. The Mobley order suggests plaintiffs no longer need to route every claim through the employer to get at the technology itself, and that a vendor’s home jurisdiction can follow its software wherever it is deployed.

The case has also been building toward this point for over a year. A separate order in May 2025 granted conditional certification of an age-discrimination class covering applicants 40 and older who were screened by Workday’s system going back to September 2020, before the FEHA ruling. Taken together, the two orders point toward a case that could eventually cover a very large pool of rejected applicants across many employers, not a single company’s hiring decision.

What it means for the HR leader

For HR and talent acquisition leaders, the practical takeaway is not about Workday specifically. It is that the assumption “our vendor’s terms of service protect us” is getting tested in real time, in both directions: employers may find vendors more exposed than expected, but that does not mean employer liability shrinks.

Newsletter

Get the week's best tech coverage.

Free. Read by thousands of HR, tech, and business leaders.

Three moves are worth making now, ahead of any trial outcome. First, ask AI screening and recommendation vendors directly what bias-audit documentation they can produce, and how often it is refreshed, rather than accepting a general compliance statement. Second, revisit indemnification language in existing contracts specifically for algorithmic discrimination claims, since many agreements were drafted before this theory of vendor liability existed. Third, keep a human review step, and a record of it, in any AI-assisted screening or ranking decision, since the strength of a proxy-discrimination claim often turns on whether a human could have caught what the algorithm did.

None of this requires abandoning AI-assisted hiring tools. It requires treating vendor selection and contract review as a compliance function, not just a procurement one, while a case that could reset the rules for the entire recruiting-technology industry works its way toward trial.

The population this could eventually cover

The scale question is what makes this case different from a typical single-plaintiff discrimination suit. The 2025 conditional class certification covers applicants 40 and older who were evaluated by Workday’s screening system going back to September 2020, a multi-year window across a platform used by employers well beyond California. The FEHA ruling adds a second legal track on top of that age-discrimination class, one built on disability and, potentially on amendment, other protected characteristics. Even with two claims dismissed on procedural grounds this round, both plaintiffs retain the chance to refile with the court’s permission, which is why litigation watchers are treating the order as an expansion of the case’s reach rather than a narrowing of it.

That combination, a certified age class plus a surviving state-law theory that follows the vendor’s home jurisdiction, is what other AI vendors and their enterprise customers are watching closely. A ruling against Workday would not bind other companies as a matter of law, but the reasoning would be available to any plaintiff’s attorney building a similar case against any other recruiting-technology platform designed or operated out of California, which covers a large share of the sector.

Source: U.S. District Court, Northern District of California (Mobley v. Workday order, June 22, 2026)