LIVE FEED // CLASSIFIED INTAKE // UPDATED Nov 20, 2026 12:41 EST
▸ Lawsuits · Nov 20, 2026

Washington Wants AI in Hiring. Workday’s 15,000-Applicant Lawsuit Shows the Risk

Workday wants a federal court to break up a massive age-discrimination collective while Washington tells agencies to put more artificial intelligence into hiring. The legal question is no longer abstract: who is accountable when software quietly decides who gets seen?

Author: [REDACTED] 1 reads

November 20, 2026 — The American job market looks calm in the aggregate. The machinery deciding who gets through the door is anything but.

Workday is asking a federal judge in San Francisco to decertify an age-discrimination collective involving more than 15,000 job applicants. The applicants say Workday's AI-assisted recruiting tools unlawfully screened out people 40 and older. Workday says the evidence does not show that its tools evaluated most of the applications at issue, much less caused the applicants' rejection, and argues that employers configure and use the software differently.

That filing, made November 10 in Mobley v. Workday, Inc., is a defense motion—not a ruling. The underlying claims remain allegations, and the court has not decided that Workday discriminated against anyone. But the scale of the dispute makes it one of the most consequential tests yet of who can be held responsible when automated hiring systems influence access to work.

One applicant, more than 100 rejections, and a nationwide case

The lawsuit began with Derek Mobley, a Black applicant over 40 who has a disability. He alleges that he applied to more than 100 jobs at companies using Workday systems and was repeatedly rejected. The case argues that Workday's screening and recommendation products produced unlawful disparate impacts based on age, race and disability. Workday has denied the allegations and says its recruiting AI does not make hiring decisions or automatically reject applicants.

In May 2025, Judge Rita F. Lin preliminarily certified an opt-in collective under the Age Discrimination in Employment Act. Later amendments added plaintiffs and claims, including claims under California law. According to current case trackers and filings, plaintiffs moved for class certification in September 2026, Workday filed its opposition and decertification bid on November 10, and a hearing is scheduled for March 9, 2027.

The fight now turns on a question that technology companies prefer to blur: is the software merely a neutral tool used by employers, or can the vendor itself function as an employer's agent when its system scores, sorts, ranks or screens applicants?

Washington is moving in the opposite direction

While thousands of private-sector applicants challenge algorithmic screening, the federal government is encouraging agencies to adopt more AI in hiring.

An August 27, 2026 memorandum from the U.S. Office of Personnel Management says agencies may be compromising hiring efficiency and quality by failing to use AI. It directs agencies to leverage AI while following federal governance requirements and pursuing the government's target of reducing average time-to-hire to 80 days.

CBS News reported in September that AI interviews were being introduced for some federal hiring, beginning with the Tech Force program. The tools can conduct early-stage interviews through virtual agents, phone, audio or text. That may accelerate hiring. It also raises the same accountability problem now before the court in the Workday case: if an automated system disadvantages a protected group, who can explain the decision, audit the model and make the applicant whole?

The government cannot credibly demand transparency from private employers while building a federal hiring pipeline that applicants cannot inspect. Faster processing is not the same as fair processing.

A strong labor market can still hide a broken application process

Fresh national data gives officials a reason to sound optimistic. The Bureau of Labor Statistics reported that employers added 165,000 jobs in October and unemployment held near 4.1%. Initial unemployment claims rose only slightly to 203,000 in the latest week and remain historically low.

But national averages do not reveal what happens inside an applicant tracking system. A worker can submit dozens of applications, receive instant form rejections and never learn whether a recruiter saw the résumé. New reporting this week found that applicants increasingly rewrite files to get past AI filters because the speed of rejection suggests no human review occurred.

This is the modern labor-market contradiction: the government counts openings and hires, while applicants confront invisible scoring systems whose criteria may be proprietary, employer-specific or impossible to challenge.

Workday's broader pressure campaign

Workday is simultaneously selling an expansive vision of AI agents in the workplace. At Workday Rising EMEA this week, company leadership described a future in which people ask and agents act—chasing approvals, connecting data and handling routine work. On November 18, Workday also announced plans to acquire Honeydew, a semantic-layer company intended to improve AI and business-intelligence access to live operational data.

That product push arrives amid business pressure. Workday disclosed another workforce reduction in September, cutting roughly 2.5% of its staff and estimating related charges of $65 million to $80 million. Earlier cuts and management changes have added to questions about growth, government sales cycles and the company's effort to reposition itself around AI.

None of those facts proves the allegations in Mobley. They do, however, show the conflict at the heart of the case: Workday is asking customers and investors to trust AI with more consequential work while asking a court to narrow responsibility for alleged harms tied to AI-assisted hiring.

The standard should be simple

If a system can influence whether a human being gets considered for a job, that system must be auditable. Applicants should receive meaningful notice when automated tools are used. Employers and vendors should retain records sufficient to test disparate impact. And responsibility should follow actual control—not marketing language that calls a consequential screening system a mere recommendation engine.

The Workday lawsuit may determine how far existing civil-rights laws can reach into the software layer of employment. The federal government's own expansion of AI hiring makes the answer urgent. A black box should not become a legal dead zone simply because it processes applications faster than a human can.

Sources

Editor's note: Litigation claims described above are allegations. Workday denies wrongdoing, and no court has ruled on the merits that its products discriminated against applicants.