California sues over AI in hiring and pricing as algorithm accountability enters court

A California court will hear claims that Workday's AI-powered job screening tools discriminate against older applicants and people with disabilities. The lawsuit alleges that the company's algorithms, which filter job candidates before a human ever sees a résumé, systematically exclude protected groups. Workday has not disclosed the training data or decision weights behind its tools, a pattern common to most vendor-provided hiring systems. The case survived an early motion to dismiss.

The same week, California filed a separate suit against BP, Marathon, 7-Eleven, and Walmart, alleging that AI-driven pricing algorithms enabled coordination that inflated gas prices across the state. Both cases share a legal architecture: plaintiffs are asking courts to pierce the proprietary screen around algorithmic systems and require disclosure of how decisions are made and whose outcomes suffer.

For hiring, the consequences are occupational. Junior roles in customer success, paralegal work, and data analysis have moved almost entirely to automated first-pass screening in the past three years. The efficiency is real—one enterprise recruiter told me her team reviewed 80 percent fewer résumés after deploying Workday. But no one tracks what happens to the 80 percent, and whether the cuts fall evenly. The lawsuit argues they do not.

If the court orders Workday to open the model, every HR vendor will face the same pressure. The sector has operated on a promise: black-box tools save time and reduce bias because they strip out human prejudice. The California complaints argue the opposite—that learned bias is harder to audit, and that scale makes the harm structural rather than anecdotal.

This is not an academic debate anymore. The tools are in production. The suits are in discovery. And the workers who never see the rejection email are starting to ask what happened before the email was never sent.