A federal judge has rebuffed Workday's argument that it cannot be held liable under California's anti-discrimination law when its AI tools screen job candidates in other states. US District Judge Rita Lin indicated this week she will likely allow additional state discrimination claims against Workday to move forward, expanding a case that could hold HR software vendors directly responsible for bias embedded in algorithmic hiring-even if they aren't the final employer.
The expansion of liability under California law
The dispute centers on California's Fair Employment and Housing Act (FEHA), one of the strongest anti-discrimination statutes in the country. Workday had asked the court to dismiss claims brought under FEHA, arguing the law should not apply to out-of-state employers or job seekers simply because the company uses its platform. Judge Lin disagreed, saying FEHA does apply and that Workday can be directly liable for "its own engagement in FEHA-regulated activities on the employer's behalf."
Holding businesses liable for their own discriminatory conduct, she noted, is consistent with the law's purpose and public policy. The ruling, once finalized, could set a precedent that any AI tool materially influencing who gets hired or rejected falls under state anti-discrimination enforcement, regardless of where the employer is based.
How the case against Workday took shape
The lawsuit, Mobley v. Workday, Inc., was filed in 2024 by Derek Mobley, a Black disabled man over 40 who applied for more than 100 positions through Workday's platform. He alleges the company's automated screening algorithms repeatedly rejected him based on race, age, and disability. The claims cite both federal statutes-Title VII, the ADA, and the ADEA-and California's FEHA.
Plaintiffs argue that AI-driven screening tools can reproduce bias even when protected characteristics aren't explicitly fed into the model. Training data, model design, and evaluation criteria can all introduce discrimination. For example, years of experience may serve as a proxy for age, employment gaps may correlate with disability or caregiving duties, and educational background may reflect race.
Federal judges previously allowed key parts of the lawsuit to proceed, ruling Workday could be treated as an "agent" of the employer under anti-discrimination law. The latest decision signals that FEHA claims will also move forward, widening the legal exposure for ai-powered hiring tools (those looking to build deeper expertise in this area can explore AI for Human Resources resources that cover bias testing and compliance).
Workday's defense rests on human oversight
Workday has denied the allegations. A company spokesperson said the claims are "false," and that "Workday's AI recruiting tools don't make hiring decisions and are designed with human oversight at their core. Our technology looks only at job qualifications, not protected traits like race, age, or disability. We rigorously test our products as part of our responsible AI program to confirm our tools do not harm protected groups."
Kelly Trindel, Workday's Chief Responsible AI Officer and former chief analyst at the EEOC, leads a team of psychologists and data scientists focused on ensuring the company's AI is "responsible, fair, and ethical." She said the AI does not make employment decisions or determine who gets a job, and that "Workday builds AI to support people, not replace them, and this is of particular importance when it comes to hiring." Trindel also pointed to Workday's independently evaluated AI governance program based on NIST and ISO standards.
What the ruling means for AI hiring tools
Workday isn't the only vendor under legal scrutiny. Eightfold faces a separate California class action alleging its tools unfairly use online data to predict candidate fit. The mounting cases put pressure on enterprises to vet AI recruiting systems more carefully. Valence Howden, advisory fellow at Info-Tech Research Group, said, "This case reinforces the importance of actually managing AI risks. If an AI-enabled model or ATS is making decisions based on historical information, it can raise questions about whether bias in outcomes and datasets has been properly addressed."
Howden warned that many organizations still treat bias validation as a one-time exercise rather than an ongoing process. "Validation of non-biased outcomes also needs to be active and ongoing, rather than a point-in-time exercise," he said. HR leaders who supervise the procurement and use of these platforms might consider a structured AI Learning Path for HR Managers to build the evaluation skills that these legal developments demand.
The risk, he added, is that AI can mimic past discriminatory decisions and penalize different communication styles. "It's easy to forget that AI can emulate and adapt those biases as part of its perspective," Howden said. Even if human oversight is stated as a core principle, "it's hard to incorporate humans into the process if the platform does the weeding out before humans have the ability to intervene."
Why this matters for HR professionals
The case makes clear that legal liability for biased AI hiring isn't confined to the employer. Vendors may be held accountable, but that doesn't reduce an organization's burden. HR departments should be actively defining how screening tools operate inside their hiring workflows, conducting ongoing bias audits, and documenting the logic behind every automated decision. A point-in-time check of an algorithm's fairness won't hold up if the model learns from biased data over time.
Procurement teams need to ask hard questions about training data, proxy variables, and how "relevance" scores are computed. The scrutiny extends beyond the software itself to the entire decision-making process-who intervenes, when, and with what evidence. As the courts draw clearer lines, HR leaders can't afford to treat AI risk as a compliance checkbox. It demands continuous validation, transparent governance, and a workforce equipped to spot when the tool, not the candidate, is creating barriers.
Your membership also unlocks:
