Workday asks a federal judge to break up the AI hiring bias case against it - Startup Fortune

Workday is pressing a federal judge to dismantle the sprawling group lawsuit that accuses its AI-powered hiring tools of systematically discriminating against older, minority, female, and disabled job seekers — a move that, whatever the outcome, is set to define the legal ground rules for every vendor selling algorithmic screening software.

On November 10, Workday filed two motions in the U.S. District Court for the Northern District of California. The first asks Judge Rita Lin to decertify a collective of more than 15,000 opt-in plaintiffs pursuing age-discrimination claims under federal law. The second asks her to block a potentially far larger group — which Workday says could number in the millions — from jointly pursuing race, sex, age, and disability discrimination claims. The filings were reported by MLex and Forbes.

Neither motion is a ruling. Lin has scheduled a class-certification hearing for March 2027, and neither she nor the plaintiffs' lawyers had responded publicly to the new filings at the time of publication.

**The core defense** Workday's argument rests on two pillars. First, it contends that hiring decisions belong to employers, not to Workday, and that the named plaintiffs have not produced evidence to the contrary. Second, it says that because each of its more than 11,500 customers configures the software differently, each rejected applicant's claim would require its own individual review — the opposite of what a class or collective action is designed to handle.

According to MLex, Workday's filing stated: "Eighteen months after preliminary certification, Plaintiffs have a heavier burden: they must prove, by substantial evidence, that the 15,000+ opt-in Plaintiffs are similarly situated in ways that are material to their claims." The company is seeking dismissal without prejudice, meaning individual plaintiffs could still bring their own suits.

In a statement quoted in coverage of the filing, Workday said its technology "focuses on qualifications and is tested for potential bias."

**How the case reached this point** Derek Mobley, a Black applicant over 40 who has disclosed anxiety and depression, filed suit in 2023 after being rejected from more than 100 positions at companies using Workday's tools. His theory — that the software's own scoring constitutes a discriminatory policy — has survived several attempts by Workday to narrow the case.

In May 2025, Judge Lin allowed the age claim to proceed as a collective, finding that Mobley had substantially alleged a unified policy: Workday's AI recommendation system scoring, sorting, ranking, or screening applicants. That was a threshold finding, not a ruling on discrimination. In February 2026, Lin approved notice to anyone who applied through Workday since September 24, 2020 and was 40 or older at the time — a pool Workday had warned could reach hundreds of millions. Lin reportedly answered that "allegedly widespread discrimination is not a basis for denying notice."

Plaintiff filings put the total number of applications rejected through Workday's tools in the relevant period at approximately 1.1 billion, though no court has found that any discrimination actually occurred.

**What hangs on the March 2027 hearing** The decertification stage is widely regarded as the harder fight for plaintiffs. Workday now gets to argue that the evidence gathered since preliminary certification does not hold up under a more demanding standard.

For the HR-tech industry, the most consequential scenario may not be a clean win or loss on the merits. If Lin keeps even a narrower group alive, the discovery process — which would force Workday to explain how its scoring features work — continues for another year or more, exposing trade-sensitive algorithmic details. If the groups are decertified entirely, individual suits remain possible, but the economics of pursuing them are vastly different.

Workday's vendor-responsibility defense is also the one every AI recruiting startup is likely to invoke. A ruling on it will carry well beyond this single docket, shaping how courts evaluate liability across the screening-software market.

For employers using automated hiring tools, the plaintiffs' theory carries a direct implication: if the software's scoring is itself the alleged policy, a vendor contract does not automatically absorb the legal risk. Procurement teams and HR leaders would do well to ask vendors now what bias-testing has been conducted, who holds the results, and what documentation is available for review.