cd /news/ai-policy/workday-asks-a-federal-judge-to-brea… · home › topics › ai-policy › article
[ARTICLE · art-168104] src=startupfortune.com ↗ pub= topic=ai-policy verified=true sentiment=· neutral

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

Workday filed papers on November 10 in federal court in San Francisco asking Judge Rita Lin to decertify the age-discrimination collective in Mobley v. Workday, which now covers more than 15,000 opted-in applicants, and in a separate filing asked Lin to block a wider group Workday says runs into the millions from pursuing race, sex, age and disability claims together, according to MLex and Forbes. No ruling has been issued; Workday argues the named plaintiffs have not shown common evidence tying the classes together and that employers configure the software differently, contending that "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 outcome will determine how much shared discrimination liability AI hiring vendors carry, following Lin's May 16, 2025 order letting the age claim proceed as a collective and her February 2026 approval of notice to applicants 40 and older who applied through Workday since September 24, 2020.

by read5 min views3 publishedNov 12, 2026
Workday asks a federal judge to break up the AI hiring bias case against it
Image: Startupfortune (auto-discovered)

No judge has ruled that millions of applicants can't sue Workday together. The company is asking for exactly that, and the answer will tell every AI hiring vendor how much shared legal risk it carries.

On November 10, Workday filed papers in federal court in San Francisco asking Judge Rita Lin to decertify the age-discrimination collective in Mobley v. Workday, which now covers more than 15,000 people who opted in, according to MLex and Forbes. In a separate filing, it asked her to block the wider group of applicants, which it says runs into the millions, from pursuing race, sex, age and disability claims together. Some headlines this week made that read like a decision. It isn't one. It's a request, and Lin hasn't ruled on it.

The distinction matters if you sell or buy screening software. A court order ending the group case would be a real win for vendors. A motion is only a statement of how hard Workday plans to fight.

The company's main line is that it doesn't make hiring decisions, its customers do. Workday says the named plaintiffs haven't shown otherwise, which leaves no common evidence to hold the proposed classes together. It also says employers configure the software differently, so each applicant's claim would need its own review. Its filing put the burden plainly: "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," MLex reported. Workday wants the claims dismissed without prejudice, so applicants could file individual cases.

In a statement quoted in coverage of the filing, Workday said its technology "focuses on qualifications and is tested for potential bias," and noted that more than 11,500 organizations use it.

A federal judge's ruling against Workday puts every AI hiring vendor on notice for discrimination liability

A California federal judge ruled on June 16, 2026 that Workday can face liability under state civil rights law for AI-driven hiring discrimination, even for employers operating outside California. The case, brought by Derek Mobley and covering 1.1 billion rejected applications, establishes an 'agent' theory of vendor liability that puts every... - AI hiring discrimination liability Workday - algorithmic age discrimination screening software

The timing is deliberate. Workday's motion came 543 days after Lin's May 16, 2025 order let the age claim proceed as a collective, when she found Derek Mobley had "substantially alleged the existence of a unified policy: the use of Workday's AI recommendation system to score, sort, rank, or screen applicants." That first stage was a low bar. The decertification stage is the hard one, and Workday gets to argue that the evidence gathered since doesn't hold up.

How the case got here #

Mobley sued in 2023. He is a Black applicant over 40 who says he has anxiety and depression and was rejected from more than 100 jobs that used Workday's tools. Lin has been unfriendly to Workday's attempts to narrow the case so far. In February 2026 she approved notice to anyone who applied through Workday since September 24, 2020 and was 40 or older at the time. Workday had warned the group could reach "hundreds of millions" of people, and she answered that "allegedly widespread discrimination is not a basis for denying notice," as Law360 and other outlets reported.

In June, Reuters reported that Lin also refused to throw out a claim that Workday's software flags applicants using proxy indicators of disability, in violation of the Americans with Disabilities Act, and let claims under California law go forward. Court filings in the case put the number of applications rejected through Workday's tools in the relevant period at about 1.1 billion. Nobody has yet ruled that the tools discriminated. That question is untouched.

Lin has scheduled a class-certification hearing for March 2027 to take up these motions.

What it means if you build or buy screening tools #

Frankly, the outcome that should worry vendors isn't a loss on the merits. It's the middle path, where Lin keeps a narrower group alive and the discovery fight over how the scoring features work keeps running for another year. Workday's defense that employers make the final call is also the one every AI recruiting startup will reach for, so a ruling on it will travel well beyond this docket.

If Workday wins, individual suits can still go ahead, and each one needs its own evidence. If it loses, the biggest group claim against an AI screener stays intact. Either way, an employer that runs automated screening can't assume the vendor's contract absorbs the risk, since the plaintiffs' theory is that the software's scoring itself is the policy. Ask your vendor now what bias testing it has run, who holds the results, and what you can see of them. Neither Lin nor the plaintiffs' lawyers have responded publicly to the new filings, and the court has not set a date for a ruling before the March hearing.

Meta Weighs Cutting 16,000 Jobs to Fund Its AI Infrastructure Bet Reuters reports the cuts would hit non-AI teams while Meta's 2026 capital spending climbs to as much as $135 billion, nearly double last year's outlay, with a $600 billion data center build-out planned through 2028. - Meta layoffs to fund AI infrastructure investments - tech companies cutting staff for artificial intelligence spending

Also read: Databricks launches always-on Genie agents to take on Anthropic and Microsoft • BloombergNEF finds AI data centres plan 181GW of their own gas power • Google's Gemini Decision 1 destroys the so-called system one models

This article is posted in AI News, check it out for more related stories.

Join the discussion #

Open in the community → Almost there. Sign in and your reply posts straight away.

── more in #ai-policy 4 stories · sorted by recency
── more on @workday 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/workday-asks-a-feder…] indexed:0 read:5min 2026-11-12 · —