Over the course of several years, Derek Mobley applied for over 100 jobs. Every time, he was turned down. The rejection itself wasn’t what made his experience unique; rejection is commonplace in the harsh job market. The speed was particularly noteworthy. Notifications were returned in a matter of hours. Occasionally, overnight. Before a reasonable person could have finished reading his resume, a few showed up.
Mobley, a Black man over 40, began to suspect that no human decision-makers were involved at all. In 2023, he filed a federal lawsuit against Workday, Inc., the software company whose AI-powered tools were screening his applications long before a human recruiter ever saw his name, rather than the specific employers who rejected him.
That distinction is very important. Additionally, a Northern California federal judge concurred that it was worth investigating.
In this market, Workday is a major player. Since its founding in 2005, the company has expanded to 175 countries and now provides services to over 65% of the Fortune 500. Thousands of organizations use its enterprise software to manage their finances and human resources. Because the AI hiring tools at the heart of this case are integrated throughout that entire infrastructure, the potential impact of any discriminatory pattern is not restricted to a single industry or employer. It’s almost intentional, systemic.
In 2024, U.S. District Judge Rita F. Lin rejected Workday’s motion to dismiss, stating that the business might be held accountable as a “agent” of the employers utilizing its platform. That was noteworthy enough. However, things went much further with the May 2025 ruling that granted preliminary certification for a nationwide collective action. Mobley’s Age Discrimination in Employment Act claim can now proceed as a class action, enabling employees 40 years of age and older who have been turned down via Workday’s platform since September 2020 to participate.
Workday speculated that the class might have hundreds of millions of members and contended that identifying collective members would be an administrative burden too big to handle. Judge Lin gave a sharp reply. If the number of impacted applicants was really that high, she wrote, it was because Workday had been credibly accused of discriminating against a huge number of people. However, this scale by itself did not justify denying potential plaintiffs notice.
Something about that logic is almost illuminating. The more people a system affects, the larger it is. Controlling the legal exposure gets more difficult the more people it touches. Workday’s legal team may have recognized this from the start and hoped that the magnitude of the issue would make it unfeasible to pursue. That calculation seems to have backfired, if that is the case.
Workday publicly asserts that customers maintain complete control over hiring decisions, with human oversight throughout, and that its AI tools do not make such decisions. According to Kelly Trindel, the company’s chief responsibility officer, the system is not made to automatically reject applicants. How that fits with Mobley’s story of overnight rejections of over 100 applications is still up for debate. Either something else was going on upstream, or the humans reviewing those applications were incredibly quick.
There is more to this case than just Workday. HR specialists and employment attorneys observing from the sidelines are aware of the potential consequences for all employers using third-party AI screening systems in the event of a decision against the company. A precedent establishing vendor liability would compel businesses to audit the systems they license as well as their own operations. It’s a big change in operations, and many organizations might not be ready for it.
There is still a long legal road ahead. After additional discovery, Workday will have the chance to contest the class certification, possibly claiming that plaintiffs are not sufficiently similar to proceed collectively. However, the door is now wide open. For employees over 40 who feel that their fate was secretly decided by an algorithm before a human ever looked at their file, that could be the first genuine confirmation that their suspicions should be taken seriously.
