After Mobley v. Workday: How HR Should Audit AI Hiring Tools
"The vendor said it was fair" is not a hiring policy.
That's the lesson from Mobley v. Workday, the lawsuit claiming Workday's AI screening tools rejected applicants based on age, race, and disability. On Inclusive AF, Katee Van Horn and Jackye Clayton broke down what the case means for anyone running a hiring funnel through software they didn't build and can't fully see.
Where the case stands
In 2025, the court let a nationwide age discrimination collective move forward. In June 2026, Judge Rita Lin denied Workday's motion to dismiss the plaintiffs' California FEHA claims, holding that Workday can be directly liable for its own screening conduct performed on employers' behalf. An ADA claim based on proxy discrimination also survived. A newly added race claim was dismissed on procedural grounds.
Translation: the vendor can be on the hook. That doesn't take you off it.
The tool doesn't own your outcomes. You do.
AI screening speeds up the part of hiring where bias is hardest to see: who never reaches a human. If a model learned from past hiring decisions, it learned whatever patterns were in those decisions. Nobody on your team approved those patterns. They're still your results.
What to check in your own funnel
- Pass-through rates by stage. Applied, screened, interviewed, offered, hired. Find where groups drop off and ask why.
- Knockout questions and filters. Graduation dates, employment gaps, and location filters can stand in for age, disability, and race.
- Who can override the score. If recruiters can't see why a candidate ranked low or can't overrule it, the tool is making the decision.
- Your disclosures. NYC Local Law 144 requires bias audits and candidate notice for automated employment decision tools. Other states and the EU AI Act are adding their own rules.
Questions to ask your ATS vendor
- How does the system detect or flag potential bias in screening or selection?
- When was your last bias audit, who ran it, and can we see the results?
- Can you explain how a candidate's match score or ranking is calculated?
- Can recruiters override or contest the system's recommendations?
- Can we hide demographic details and identifying information during screening?
- Can we track pass-through rates by demographic group at every stage?
- How do you support compliance with EEOC guidance, NYC Local Law 144, and the EU AI Act?
If the answer to most of these is "we're working on it," you have your answer.
The bottom line
Buying AI for hiring is a decision about who gets considered for work at your company. Treat it like one. Audit before you buy, audit after you launch, and keep a person in the process who can explain every rejection.
Listen to the full episode with Katee and Jackye for the stories behind it.
Co-Host
Jackye Clayton has spent more than 15 years in HR and HR tech, most of it arguing that hiring is a design problem, not a pipeline problem.
She led talent and inclusion work at Textio and SeekOut and edited at RecruitingDaily. Today she runs Talent Lab, advising companies on AI in hiring, neurodiversity, and talent strategy. She founded People Puzzles, teaches on LinkedIn Learning, and reports from the field at conferences like HR Tech and Workhuman Live.
She's building Who Do I Work For?, a tool that uses public records to show job seekers what a company actually does before they take the offer.
Her work has been cited in the Wall Street Journal, Forbes, Fortune, and SHRM. She's on HR Executive's Top 100 HR Influencers list.
She lives in Waco, Texas, where she serves as Secretary on the board of Creative Waco and Chair of the board of Communities In Schools of the Heart of Texas.