Derek Mobley applied for more than 100 jobs at companies using the same AI-enabled screening platform. He says he was rejected every single time. Some of those rejections, according to his lawsuit, showed up within the hour, in the middle of the night, when no human recruiter was awake to read anything.

Mobley is over 40. His case against Workday has become the most closely watched AI hiring lawsuit in the country, and it rests on the idea that the software never needed to know his age. It could infer it.

An Old Problem

Age bias in hiring isn’t new. AARP’s research consistently finds that roughly two-thirds of workers 50-plus have seen or experienced age discrimination at work, and three in four believe their age will be a barrier to getting hired. The most common thing they report? The assumption that older workers aren’t tech-savvy.

What’s new is the delivery mechanism. We’ve taken that old assumption and automated it.

An AI screener trained on “who we hired last time” learns to reward the patterns of your past hires, and it has plenty of age proxies to work with: graduation year, length of work history, “vintage” tools listed on a resume.  (If your AI tool has learned to downrank anyone still using a Hotmail address, congratulations. You’ve built a very expensive age detector.)

Then layer on the process itself. A timed one-way video interview. A gamified assessment piloted on a room full of 24-year-olds. A chatbot-only application with no human path. A job description asking for a “digital native” or capping experience at “3–5 years.” Each one looks like efficiency. Together, they’re a funnel that leaks out the age demographic, silently and at scale.

Would you even notice? Most organizations don’t collect applicant age, which means most Applicant Tracking System (ATS) reports can’t show you this leak even if you were looking.

Your screening tool doesn’t need to know a candidate’s age. A graduation year and a 25-year work history tell it everything

The Loyalty Factor

This is more than a compliance issue. It’s a retention issue hidden in the sourcing strategy.

The most recent Bureau of Labor Statistics (BLS) data puts median tenure for workers 55 to 64 at 9.6 years, more than three times the 3.0 years for workers 25 to 34. A joint LinkedIn and AARP analysis found that 85% of workers 50-plus hired in mid-2024 were still with their employer a year later, versus about 71% of younger hires.

The “non-tech savvy” story also doesn’t stick. The same LinkedIn and AARP report found the number of workers 50-plus listing AI among their skills grew 25% over five years, nearly double the growth rate of younger workers. Meanwhile, only about one in eight older workers has received any AI training at work, and roughly half say they want it. That’s not a capability gap. That’s an access gap that employers have built.

So the candidate your tool flagged as a weak fit for “our fast-paced environment” may well be the one still on your team after your next three hires have moved on.

Screening out experienced candidates isn’t just an inclusion problem. It’s bad retention strategy.

The Legal Case

Mobley is no longer a fringe case. A federal court conditionally certified a nationwide age-discrimination collective in May 2025, and in March 2026 rejected Workday’s argument that the ADEA’s disparate-impact protections don’t reach job applicants. Industry reporting puts opt-ins around 14,000, and Workday’s own court disclosures showed its tools had rejected applications numbering in the billions during the relevant period. The court has also treated the AI vendor as a potential agent of the employers using it.  Translation:  “we bought it from a vendor” is not a defense that holds any ground.

Note:  Talk to employment counsel about any exposure at your specific company. I’m a recruiting guy, not a lawyer, and this is still very much a live case.

The Fix

None of this requires ripping out your technology. It requires asking better questions about how it’s configured.

Understand if the role needs to know when someone earned a degree or only that they did earn a degree.  Suppress dates from AI scoring.

Ask questions about why experience is capped.  Is it about skill or budget?  If it’s budget, post the range and let candidates self-select. I made the broader case for this in my skills-first job description series Part 1 and Part 2.

Don’t screen on a checklist of current technologies. Instead ask yourself (or the hiring manager) if this person can learn our stack in 30 days? If yes, it’s a training line item, not a knockout question.

Embrace the technology but also create a human path for candidates who want one. Offering a choice of format costs you almost nothing.

Instead of “digital native” or “high-energy recent grad mindset,” ask what behavior do we actually mean? Then write that.

And don’t assume your process works for everyone. Look for older and younger persons to run through it before it’s launched.

Campus Recruiting Gets It

It seems every week my writing turns to campus recruiting. Maybe it’s because I led a large campus recruiting program for several years or maybe it’s because campus recruiting gets it.

Early talent teams are the best at designing a process around its audience. They pilot applications with actual students, go mobile-first because that’s where Gen Z lives, and schedule by text because nobody under 25 answers a phone call. Nobody would let a screener knock out a graduating senior for “insufficient experience,” because the tool is calibrated to the population it serves.

Campus teams also have a legitimate reason to ask for graduation year. It’s the one place in your hiring process where that field genuinely belongs. Carry it over into experienced hiring and the same data point that powers your campus program becomes a liability.

The lesson here is to test your experienced-hire funnel the way you’d test your campus funnel. Your 55-year-old controller candidate deserves the same attention to the design of the process as your 21-year-old finance intern.

The AI Factor

I hear pushback quite a bit: “We cannot afford to hire people who cannot keep up with AI”. That’s fair and the simple fix is not to zero in on the population you think has the skill. Just test for it. A short, practical skills exercise on the tools the role actually uses will tell you far more than a model inferring comfort with technology from a birth decade. And if a strong candidate is a few weeks short on a specific platform, that’s the upskilling offer, not a rejection. If you need convincing of this, please read my recent article titled, The High Cost of Chasing Unicorns.

The other objection I hear is quieter: “They’ll just retire in a couple of years.” Run that against the tenure data above. A 58-year-old with seven good years ahead of them is likely to outlast the median 30-year-old hire twice over. That’s right, there’s less risk in hiring more experienced employees because they tend to stay much longer.

---

AI isn’t creating age bias in hiring. It’s scaling the version we already had and removing the moment where a human might have caught it. The companies that win here won’t be the ones that abandon their tools. They’ll be the ones that configured them with purpose, audited them honestly, and kept a door open for the candidates their algorithms don’t know how to read.

When was the last time your team checked whether your screener treats a 25-year career differently from a 5-year one? Drop a comment. I’d genuinely like to know how many teams have actually looked.

---

If you want a second set of eyes on how your ATS, assessments, and AI screening are configured before someone else goes looking, that’s exactly the kind of work ES Talent Solutions does. Reach me directly at estewart@ESTalentSolutions.com and let’s talk.

#Ageism #AIinHiring #AgeInclusion #TalentAcquisition #RecruitingStrategy #Hiring

---