Federal court orders Workday to answer claims of discrimination in AI candidate screening

A US federal court has allowed a class action against Workday to proceed. A San Francisco judge ruled that the company must answer allegations of discrimination in AI-assisted hiring. The case is being called the first to broadly challenge the algorithms of candidate-screening software.

The first lawsuit against a hiring algorithm

Historically, most discrimination claims in employment were brought against employers, who were responsible for the final decision to hire or reject. This lawsuit is built on a fundamentally different footing. The target is the vendor itself: Workday, one of the world's largest developers of enterprise HR systems, with hundreds of thousands of corporate customers.

According to the plaintiffs, Workday's algorithms created discriminatory barriers before any human recruiter had a chance to look at a résumé. The mechanism worked as an invisible filter: the system assigned candidates a score and decided whether to pass them further down the funnel, with no human involved.

Lawyers and industry observers describe the case as a precedent. For the first time in the US, a class action directly and broadly targets not a specific employer but the mechanics of an AI hiring tool itself, and the company that built it and sold it to thousands of organizations.

What Workday is accused of

The plaintiffs argue that Workday's algorithm filtered candidates at client companies in ways that violate California law. The details of the claims have not been fully disclosed, but the allegations reportedly concern systemic discrimination based on age, race or disability.

Key elements of the complaint:

• The AI algorithm automatically scored candidates and screened some of them out before a recruiter was involved
• Rejections came with no explanation, so candidates could not challenge the machine's decision
• Workday's client companies may not have been aware of the bias built into the model
• The same algorithm was used by thousands of employers, which substantially widens the scale of potential discrimination
• According to the plaintiffs, California anti-discrimination law applies to the software developer, not only to the end employer

Workday denies wrongdoing. The company has already tried to get the lawsuit dismissed, but the judge found the plaintiffs' arguments strong enough for the case to go to full consideration.

Why this changes the rules

Until now, responsibility for discriminatory hiring defaulted to the company that made the choice to hire or not. The software vendor stayed in the background: its product was considered a tool, and the employer made the decision. This lawsuit breaks that logic.

If the court sides with the plaintiffs, developers of AI hiring systems will face a fundamentally new level of liability. They will have to prove that their algorithms contain no hidden bias, or bear the same legal consequences as discriminating employers.

For Workday, the potential consequences are especially serious. Thousands of companies worldwide use its platform, and if the claim succeeds, the scale of the claims and potential payouts could be incomparably higher than in typical hiring-discrimination cases.

What this means

The San Francisco court is sending a clear signal: a hiring algorithm is not a neutral intermediary, and its creator is responsible for how the model makes decisions. For enterprise AI developers, this means a fundamentally new regulatory risk: bias built into a model can now become grounds for a multimillion-dollar lawsuit, regardless of whether the vendor itself knew about it.