Quick Answer: Derek Mobley, a Black IT professional over 40 with anxiety and depression, applied to more than 100 jobs through employers using Workday between 2017 and 2023. Every application was rejected — sometimes within minutes, occasionally in the middle of the night. In February 2023 he sued Workday in the US District Court for the Northern District of California. In July 2024, Judge Rita Lin held that an AI screening vendor can be held directly liable as the employer's "agent" under federal anti-discrimination law. In May 2025 the same court granted collective-action certification under the Age Discrimination in Employment Act, allowing every US applicant 40 or older rejected through a Workday AI tool since September 2020 to opt in. Workday's own filings disclose roughly 1.1 billion applications processed and rejected by its AI screening tools during the covered period — making this the largest AI-hiring class action in US history. The merits are still ahead. The legal architecture, however, is now built.
Who Is Derek Mobley
The plaintiff in Mobley v. Workday is a single person whose application history reads like a stress test for the modern hiring funnel. Derek Mobley, in his fifties, holds a Bachelor of Science in Finance from Morehouse College, an associate's degree in network systems administration, and the CompTIA A+ certification. He had been working in IT and customer-service roles in the United States for over a decade. Beginning in 2017, he started applying — at first occasionally, then systematically — to roles posted through employers that used Workday's recruiting platform. By the time he filed suit, he had submitted more than 100 applications. Every one was rejected.
The pattern Mobley documented is what catches the eye. Some rejections arrived within an hour of application. One landed at 1:50 a.m., just minutes after he applied. The speed mattered legally, because no human recruiter was plausibly reading résumés at that hour. Whatever was deciding "no" was automated — and whatever it was, it was deciding "no" with extraordinary consistency for a candidate whose qualifications, on paper, fit the postings he was applying to.
His complaint, filed in February 2023 by the law firm Outten & Golden, alleges that Workday's screening tools systematically disadvantage applicants on three protected bases simultaneously: race (Mobley is Black), age (he was over 40 throughout the period), and disability (he has documented anxiety and depression). The complaint frames Workday not as a neutral software supplier but as the active screener — the actor whose algorithm, trained on whose data, made the call.
Why Workday Specifically
Workday is not the only ATS in the market. Greenhouse, Lever, Taleo (Oracle), iCIMS, and SAP SuccessFactors all process tens or hundreds of millions of applications a year. Why is the landmark case named after Workday rather than one of them?
Three reasons converge. First, scale. Workday Recruiting is dominant in enterprise: in the first half of 2024 alone, customers on its platform processed 173 million applications, up 31% year on year, against just 19 million open requisitions. That ratio — applications growing four times faster than openings — sets up exactly the conditions where automated filtering does the heavy lifting. Second, consolidation: Workday markets a match score, candidate ranking, and AI-assisted recommendations as part of its standard product, which makes the case for vendor agency easier to plead. Third, evidence: Mobley's individual rejection record across more than 100 different employers all using one common platform produced an unusually clean common variable. Different employers, same screener, same outcome.
The case is therefore not an indictment of every ATS. It is an indictment of this particular vendor's algorithm, deployed at scale, across many employers, with — the plaintiffs argue — a measurable disparate impact on protected groups. Other vendors are watching closely. EEOC charges have already been filed against Aon over AI personality assessments (ACLU, December 2023), and the regulator settled iTutorGroup for $365,000 in 2023 over an AI video-interview system that auto-rejected female applicants 55 and older and male applicants 60 and older. The legal theory in Mobley generalises; the named defendant happens to be the largest target.
The "Agent" Theory: Why a Software Vendor Is on the Hook
The most consequential ruling in Mobley arrived not at trial but at the motion-to-dismiss stage. In July 2024, Judge Rita Lin held that Workday could be sued directly as the employer's agent under Title VII, the ADEA, and the ADA. That decision is the spine of every later development in the case.
The agent theory works like this. Federal anti-discrimination statutes apply to employers and to anyone "acting in the interest of an employer, directly or indirectly." Vendors have historically argued they are mere tool suppliers, the way Microsoft is not liable when an employer uses Excel to make a discriminatory decision. Mobley punctures that defence by drawing a line: when the vendor's product materially participates in the hiring decision — by scoring, ranking, scheduling, or rejecting candidates — the vendor is acting in the employer's interest, and inherits the employer's legal exposure.
The implications travel beyond Workday. Any vendor whose product moves a candidate up or down a list, recommends rejection, or auto-routes applications now plausibly sits inside the statute. Vendors that simply store résumés are still safe. Vendors that score them are not. That is a regulatory shift the industry has spent the last fifteen years insulating itself from, and one ruling has now closed the gap.
The 1.1 Billion Applications: How Big the Class Actually Is
The May 2025 certification order is what makes the case historic in scale. The court conditionally certified a collective action under the ADEA covering every US applicant aged 40 or older who, since September 2020, was rejected from a job through a Workday AI screening tool. Workday's own discovery filings — referenced in court documents and reported by multiple legal trade publications — disclose that its AI tools processed and rejected on the order of 1.1 billion applications during the relevant period.
That number deserves a moment of arithmetic. The US labour force is roughly 167 million people. The working population aged 40 and older is roughly 80 million. If even a single-digit percentage of those workers applied to a Workday-screened job in the relevant five-year window, the over-40 subset of the 1.1 billion-application total still reaches into the high hundreds of millions of individual rejections. Estimates from class counsel run to 200 million or more potentially eligible class members. No prior US AI-discrimination case has come close to that scope.
Eligibility to opt in is narrow on its face but wide in effect: aged 40+, US-based, applied through a Workday-screened employer, was not hired. Anyone who fits is in the affected class for the ADEA claim. Class counsel's contact details are referenced through the certification order on the Civil Rights Litigation Clearinghouse, where the full case file is publicly archived under case number 44074. The Race (Title VII) and Disability (ADA) tracks remain active in the same litigation, with their own certification path still being argued.
How the Age-Discrimination Claim Actually Works
The ADEA, passed in 1967, protects workers 40 and older from age-based employment discrimination. It applies to employers with 20 or more employees, to most federal employers, and — through the agent theory Mobley just established — to vendors whose tools materially participate in employment decisions. The statutory bite is not subtle: successful claimants can recover lost wages, liquidated damages equal to lost wages where the violation is "willful," reinstatement, and attorney's fees.
For an AI-screening case, the typical legal route is disparate impact: the plaintiff does not have to prove the employer or vendor intended to discriminate. They have to show that a facially neutral practice — here, an algorithmic screening tool — produces a meaningfully worse outcome for the protected group. The Mobley plaintiffs allege that Workday's AI consistently produces lower match scores, lower interview rates, and faster rejections for applicants over 40, and that this gap cannot be explained by job-relevant qualifications. Discovery is now where the actual scoring data, the training-set demographics, and the audit history come into the open.
Workday's defence is twofold. First, they deny the disparate impact occurs at any actionable rate. Second, even where it occurs, they argue the model uses job-relevant features and that its outputs are recommendations, not decisions — leaving the human recruiter as the ultimate decisionmaker. Both defences are testable empirically, which is why the discovery phase matters so much. The certification order does not declare Workday's tools illegal. It opens the door for plaintiffs to inspect the tools, on the record, in court.
What This Sits Next To: iTutorGroup and Aon
Mobley does not stand alone. Two parallel developments give the legal climate texture.
EEOC v. iTutorGroup (2023) was the first federal AI-hiring discrimination case to actually settle. The company, which provides online tutoring, was accused by the EEOC of using AI-driven application software that automatically rejected female applicants aged 55 and older and male applicants aged 60 and older — more than 200 qualified candidates over a measurable period. iTutorGroup paid $365,000 to settle, agreed to anti-discrimination training, and accepted ongoing EEOC monitoring. The settlement is small relative to the company's size, but the precedent — that an automated cutoff rule keyed to age is a textbook ADEA violation — is fully transferable to any AI screening tool that produces similar effects without an explicit rule.
The ACLU charge against Aon, filed with the EEOC in December 2023, targets a different layer. Aon sells AI-driven personality and cognitive assessments — the type of pre-employment test deployed by hundreds of large employers. The ACLU alleges these assessments screen out neurodivergent applicants and Black applicants, in violation of Title VII and the ADA. The case is at the EEOC investigation stage, not in court, but it puts AI assessments (not just résumé screening) on the same regulatory radar.
Read together, the three cases — Mobley, iTutorGroup, Aon — cover the full pipeline: the screener, the cutoff rule, and the assessment. None has issued a final merits judgment yet. All are now on the EEOC's strategic-enforcement priority list for 2024–2028, which the agency published in 2023 alongside technical guidance on AI in hiring.
What Workday Has and Hasn't Said
Workday's public posture has been measured. The company has stated it does not make hiring decisions, that its tools surface candidates rather than reject them, and that customers retain control of cutoffs. It has announced internal bias-detection audits and committed to greater transparency. It has not released the audit results.
The gap between announced audits and published audits is itself telling. Workday is not unique here — HireVue announced and then phased out its facial-analysis tools in 2021 after academic critique, but full bias reports remain internal — but in a litigation under the agent theory, the discovery process is precisely what surfaces audit history, training-data provenance, and disparate-impact statistics that vendors have until now treated as trade secret. Mobley is the first case where that material is plausibly heading into a public docket.
What This Means for Job-Seekers Practically
Three changes are real and immediate. Three more are likely but not yet certain.
Real now: if you are 40 or older and have been rejected by a US employer using Workday since September 2020, you have a legal route to participate in Mobley. The opt-in mechanism runs through class counsel referenced in the certification order. The opt-in is free; class counsel works on contingency. Participating does not guarantee a recovery; it preserves your rights if the case settles or wins.
Real now: the EEOC has flagged AI hiring as a 2024–2028 enforcement priority and has published technical guidance describing how its existing disparate-impact framework applies to algorithmic screening. That guidance gives any individual jobseeker a clearer path to file a charge with the EEOC if they suspect discrimination — even without a class action behind them.
Real now: several US states have begun to act. New York City's Local Law 144 (effective 2023) requires employers using automated employment decision tools to conduct annual bias audits and publish summary results. Illinois requires consent and notification for AI video-interview analysis. Colorado's AI Act, signed in 2024, takes effect in 2026 and imposes risk-management duties on developers and deployers of high-risk AI systems, with employment systems explicitly named.
Likely: vendor settlements, audit disclosures, and product changes as discovery in Mobley progresses. Likely: follow-on cases against the next tier of ATS and assessment vendors using the same agent theory. Likely: a slow shift toward employers requiring bias-audit certifications from the vendors they buy from.
How to File an EEOC Charge if You Suspect Discrimination
This section is descriptive, not prescriptive. None of it is legal advice; for a specific situation, consult a US employment lawyer.
The Equal Employment Opportunity Commission accepts charges of employment discrimination at no cost. The deadline is 180 days from the alleged discriminatory act in most states, extended to 300 days in states that also have a state or local fair-employment agency. The clock runs from the rejection, not from the day you discover the cause.
The EEOC's online portal at publicportal.eeoc.gov walks through the process. You provide your contact details, the employer's contact details, the date of the alleged act, the protected basis (age, race, disability, etc.), and a description of what happened. If a vendor like Workday is identifiable from the application portal, name it; the agent-theory pleading from Mobley means the vendor can be a respondent alongside the employer. The agency reviews the charge, often offers mediation, and either pursues investigation or issues a "right to sue" letter that lets you bring your own federal lawsuit within 90 days.
Documenting the rejection is the most actionable step right now, regardless of whether you plan to file. Keep the original job description, the application timestamp, the rejection email or notification, any vendor name visible during application, and any screenshots showing automated assessment steps. Without that record, both EEOC charges and class-action opt-ins become harder to substantiate later.
The Limits — What This Case Will Not Fix Overnight
It is worth being honest about scope. Mobley v. Workday does not ban AI screening, does not declare any specific algorithm illegal, and does not entitle every rejected applicant to compensation. Class certification is a procedural milestone, not a verdict. Settlements in cases of this size typically take years, and a meaningful share of class members usually receive modest individual recoveries even when the headline number is large.
The honest read: Mobley rebuilds the legal plumbing under AI hiring. Vendors are now potentially defendants. Disparate-impact is a viable theory. EEOC and state regulators have an enforcement framework to point at. Whether any individual jobseeker's outcome improves depends on the discovery record, the settlement structure, and how aggressively the EEOC and private plaintiffs follow up after this opening case.
FAQ
What is Mobley v. Workday in plain English?
A federal class action filed in 2023 by a Black applicant over 40 who said Workday's AI screening rejected him from more than 100 jobs — sometimes within minutes — on the basis of race, age, and disability. In May 2025 the US District Court for the Northern District of California granted collective-action certification under the ADEA, opening participation to every US applicant aged 40 or older rejected by a Workday AI tool since September 2020. Workday's own filings disclose roughly 1.1 billion applications processed and rejected by its AI tools during the covered period.
Why is Workday the defendant if the employer made the hire?
Because Judge Rita Lin's July 2024 ruling held that an AI screening vendor can be sued directly as the employer's agent under Title VII, the ADEA, and the ADA when the vendor's tool materially participates in the hiring decision — by scoring, ranking, or filtering candidates. Before Mobley, vendors largely argued they were neutral software providers. The agent theory now puts the screening tool itself on the hook, not just the company that deployed it.
Does the case mean AI hiring tools are illegal?
No. The case does not ban AI hiring tools and does not declare any specific Workday product unlawful. It moves a single bias claim past the early procedural stages and into discovery. Workday denies the allegations. The merits will be decided through evidence on Workday's actual scoring data and audit history. What certification establishes is that the legal route exists, and that hundreds of millions of rejected applicants now have standing to participate.
If I think I was screened out unfairly, what can I do?
If you are 40 or older and have been rejected by a US employer using Workday since September 2020, you may be eligible to opt in to the Mobley collective action — class counsel's contact information is referenced in the certification order on the Civil Rights Litigation Clearinghouse case file (case 44074). Anyone in the US who suspects discrimination by an AI hiring tool can file a charge with the EEOC within 180 or 300 days of the rejection. Document every job description, application timestamp, rejection email, and visible vendor name. None of this is legal advice — for a specific situation, consult a US employment lawyer.
Read Next
- AI Resume Statistics 2026: 72 Verified Stats on AI Hiring, ATS, and Bias — the canonical hub with every primary source behind this article.
- AI Hiring Bias Studies 2024–2025 — the academic and litigation evidence that sits next to Mobley.
- The 75% ATS Rejection Myth, Debunked — what an ATS actually does once a résumé is uploaded, and why volume — not algorithmic gating — is the bigger filter for most jobseekers.
- JobCannon Career Match assessment — pick roles that fit your strengths before you optimise a single bullet; the easiest defence against any screening funnel is applying to roles you genuinely match.