AI Bias in Hiring is Real
On June 16, 2026, a federal judge sent a clear signal: Workday cannot hide behind state lines when its AI hiring tools discriminate. The California civil rights law may apply to every applicant its software screens, nationwide. I wrote this article to explain why that ruling matters to every career coach, interview coach, and resume writer in this profession
Imagine applying for more than 100 jobs and getting rejected every single time before a single human being ever read your name.
That is what Derek Mobley says happened to him. He is a Black man over the age of 40 with a disability who applied to more than 100 positions at companies using Workday’s AI-powered hiring platform. He got a quick rejection every time. No interview, no callback, no explanation – and not enough time between his application and rejection for human review. In February 2023, he filed a class action lawsuit that has become one of the most closely watched employment discrimination cases in the country.
If you coach job seekers, this case belongs in your awareness. Here is what it is, why it matters, and how it should shape the way you prepare clients.
The Case in Plain Language
Workday is one of the most widely deployed HR technology platforms in the country. Its AI feature called Candidate Skills Match scores and ranks applicants before a recruiter ever sees them. Mobley’s lawsuit argues the system discriminates against people who are Black, over 40, or living with a disability, in violation of Title VII of the Civil Rights Act, the Age Discrimination in Employment Act, and the Americans with Disabilities Act.
What made this case landmark was a 2024 federal court ruling that Workday itself, the software vendor, could be held liable as an agent of its employer clients. That had never happened before. Software companies can now face legal consequences for discriminatory algorithms they build and sell.
The case has kept moving. In March 2026, a judge rejected Workday’s argument that federal age discrimination law does not cover job applicants at all. The class action has been conditionally certified, meaning thousands of additional plaintiffs may now opt in. And in May 2026, the court ordered Workday to produce its EEO-1 and OFCCP compliance documents, finding them relevant to what Workday knew about demographic disparities in its own tools.
Nearly half of companies surveyed acknowledge age bias in their AI hiring tools. Some 26 percent recognize racial bias. They know. They are using the tools anyway.
Why Interview Coaches Need to Pay Attention
Yes, this is a resume screening case. But the same AI infrastructure now extends into interviews. AI-assessed video interviews ask candidates to record responses to pre-set questions. The platform then analyzes vocal tone, word choice, response structure, and content. A human may or may not ever watch the recording.
The research on what these systems actually measure is troubling. A 2025 University of Melbourne study found that AI hiring tools struggled to accurately evaluate candidates with speech disabilities or non-native accents. And in March 2025, the ACLU filed a complaint against Intuit and its AI video interview vendor on behalf of an Indigenous, deaf applicant whose video response was analyzed for active listening skills.
Let that land. An AI scored a deaf woman on active listening.
Then there is the facial expression question. Some AI hiring platforms have used technology that analyzes candidates’ facial expressions to infer emotional states and draw conclusions about their suitability for a role. HireVue was among them, until a 2019 discrimination complaint filed with the Federal Trade Commission challenged the practice as unproven, invasive, and prone to bias. HireVue dropped facial expression analysis from its platform in January 2021.”
But HireVue’s decision did not end the practice across the industry. CVS settled a case in 2024 after its AI video interviews allegedly rated candidates’ facial expressions for employability. Other tools incorporating emotion recognition technology remain available, though the regulatory landscape is tightening significantly. While the EU AI Act, in effect from February 2025, explicitly prohibits emotion recognition in workplace hiring contexts, the United States has no equivalent federal legislation. Facial expression and emotional analysis tools have not disappeared from hiring in the US.
A November 2025 University of Washington follow-up study found that human decision-makers who received AI-generated recommendations tended to mirror and amplify those biases rather than correct for them. The bias does not stay in the algorithm. It travels downstream.
What This Means for How We Coach
None of this means we tell clients the system is rigged and leave it there. It means we coach with eyes open. Here is how I think about it.
First, help clients understand that the first assessor may not be human. Especially with large employers using platforms like Workday, the initial screening of a resume or a recorded video response may be entirely automated. This is not meant to alarm them. It is meant to help them prepare for who, or what, is actually in the room.
Second, correct a coaching myth that has been circulating: modern AI interview platforms do not work by matching exact keywords. They use natural language processing sophisticated enough to understand conceptual meaning. HireVue states explicitly on its website that there are no magic keywords to trick the system, because the NLP is reading for competency, depth, and substance, not specific phrases. What this means for your clients is more interesting than keyword-stuffing advice: vague, generic answers fail with these systems for the same reason they fail with a sharp human interviewer. The AI is looking for evidence, not vocabulary.
Third, structure matters. AI assessment platforms reward responses that have a clear logical arc: context, action, outcome. That through-line is what the system can parse and score. Candidates must deliver a structured answer, but a rigid STAR-formatted answer is not necessary to satisfy the algorithm. Any response that is specific, grounded in a real example, and moves clearly from situation through action to a defined result gives the system what it is looking for. How that is organized is less important than whether all the elements are there.
Fourth, name the equity issue when it is relevant. Candidates with non-native accents, speech disabilities, or other characteristics that do not fit the narrow profile these systems were trained on face a genuine disadvantage. That is not their failure. It is a documented flaw in the technology, increasingly supported by research and litigation. Where accommodations are available, encourage clients to request them before the interview. Where they are not, help clients practice until their delivery is as clear and confident as it can be, and remind them that these systems are not the final word.
Fifth, encourage clients to track their applications. Patterns matter. The Mobley case began with one man noticing something was wrong across dozens of rejections. We are not attorneys, but we can help clients recognize patterns and know that legal protections and resources exist.
The Bigger Picture
Mobley v. Workday is not just a lawsuit. It is a data point confirming what many of us have observed for years: the algorithm is not neutral. It has absorbed the biases of its training data and is now encoding them into millions of hiring decisions at scale.
A parallel case filed in January 2026 against Eightfold AI adds another dimension, alleging the company collected and scored applicant data without consent in violation of the Fair Credit Reporting Act. New regulatory frameworks are arriving from multiple directions. The legal pressure is building, but it moves slower than the technology, and your client has an AI-assessed interview next Tuesday.
Derek Mobley applied to more than 100 jobs and never once got to talk to a human being. The case he filed may ultimately reshape how AI hiring tools are built and held accountable. That is worth knowing, worth talking about with clients, and worth folding into how we coach.
The system is imperfect and evolving. Our job is to help clients navigate it as it actually exists, with honesty, with strategy, and with the full weight of what we know.
Sources
Mobley, et al. v. Workday, Inc., Case No. 23-CV-00770 (N.D. Cal.) | Duane Morris Class Action Defense Blog, June 2, 2026 | Electronic Privacy Information Center v. HireVue FTC Complaint, 2019 | SHRM: HireVue Discontinues Facial Analysis Screening, 2021 | University of Washington (2024, 2025) | University of Melbourne (2025) | HireVue AI in Hiring Statement | ACLU Colorado / Intuit HireVue Complaint, March 2025 | Brookings Institution, April 2025 | HR Executive | AI Governance for HR | Smart Eye / Affectiva Acquisition, 2021