Workday, the $60 billion HR software company used by most of Corporate America, just lost a major bid to escape a lawsuit claiming its AI hiring tools discriminated against Black applicants, older workers, people with disabilities, and women before they ever spoke to a human recruiter.

A federal judge in San Francisco ruled Monday that the case can move forward, mostly rejecting Workday's attempts to have the amended lawsuit dismissed. It is the first lawsuit of its kind to directly target the algorithms behind AI screening software, and legal experts say it could shape how this entire category of litigation develops.

The machine in the middle

Here is how AI hiring tools generally work: a company like Workday builds software that processes thousands of job applications automatically, scoring and filtering candidates before any human sees them. The pitch to employers is efficiency. The concern from regulators and worker advocates is that if the training data used to build the software reflects historical hiring biases, the algorithm will too.

U.S. District Judge Rita Lin found that the plaintiffs had plausibly alleged exactly that. One specific claim she allowed to proceed: that Workday's software can filter out applicants based on "proxy indicators" of disability or illness, including gaps in employment history. A gap in someone's resume could reflect a cancer diagnosis, a mental health crisis, or a period of caregiving. If an algorithm quietly penalizes that gap, it may be doing indirectly what federal disability law prohibits directly.

Lin also rejected Workday's argument that California's anti-discrimination laws don't apply to people screened outside the state. Because Workday runs its operations from California, she ruled it can be held liable under state law for what its software does anywhere.

Workday pushed back hard in a statement: "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."

Why this lands on you

More than 80% of U.S. employers, and virtually every Fortune 500 company, now use AI tools in hiring, according to numerous surveys cited in the reporting. That means most job seekers today are being evaluated by software before they reach a person. Most of them don't know it.

That last part is part of why lawsuits like this have been so rare. If you never got a callback, you probably don't know whether a recruiter read your resume or a filter deleted it in seconds. The complexity of challenging algorithmic decisions, and the opacity of how these systems work, has made litigation difficult even for employment lawyers who understand what they're up against.

This case, originally filed in 2023, changes that calculus. If it proceeds to class certification and trial, it will force Workday to disclose how its models are built, what data they were trained on, and what outcomes they produce across different demographic groups. That kind of discovery could be the most consequential part of the case regardless of how it ultimately resolves.

The deeper issue is structural. AI hiring tools are sold to employers as neutral and objective, an improvement over the inconsistent judgments of individual managers. But neutrality in an algorithm is not the same as fairness. A model trained on who got hired in the past will learn, efficiently and at scale, the patterns of whoever did the hiring in the past. If those patterns were discriminatory, the algorithm preserves and accelerates them.

Federal and state agencies have flagged this problem for years. Congress has held hearings. But the legal framework for challenging algorithmic discrimination is still being built, one ruling at a time. Monday's decision is one of the first load-bearing pieces.