AI-powered HR screening technology has been operating in a legal gray zone since it entered enterprise hiring workflows. Vendors supplied the algorithms; employers made the hiring decisions; liability was distributed across a contractual chain that no plaintiff had successfully traced to its source. A federal court ruling in California on June 22 changed the terms of that distribution.
What the Mobley v. Workday Ruling Established
In Mobley v. Workday, a federal judge refused to dismiss California Fair Employment and Housing Act claims against Workday on the grounds that plaintiffs were not California residents. Judge Rita Lin found that plaintiffs had adequately tied their claims to California, where Workday is headquartered and where, the complaint alleges, the AI screening tools at the center of the lawsuit were designed, developed, maintained, and controlled. Duane Morris’ summary of the order notes the court also held that conduct inside California is not extraterritorial, regardless of where the affected applicant lives.
The ruling allows discrimination claims under FEHA to proceed against the vendor itself as well as against the employers who used Workday’s tools in their hiring processes. The lawsuit, which received preliminary collective action certification on May 16, 2025, alleges that Workday’s AI screening technology rejected job applicants based on protected characteristics including age, race, and disability. The lead plaintiff was denied employment on every one of the more than 100 applications he submitted through Workday’s platform, an outcome he attributes to the automated screening system.
The Vendor Liability Question
Prior to this ruling, the predominant legal theory in AI discrimination cases placed primary liability on the employer who deploys the technology, not the vendor who supplies it. The Mobley ruling does not eliminate employer liability, but it establishes that a vendor whose tools are designed in California is subject to California’s discrimination law regardless of where the hiring occurs or where the plaintiff is located.
The practical implication for HR technology vendors is that California’s employment discrimination framework is now effectively national in its application to any company headquartered or operating significantly within the state. Given the concentration of enterprise HR technology companies in Silicon Valley and the broader Bay Area, this jurisdictional reach applies to a significant portion of the vendors supplying AI tools to HR functions across the country.
What Changes for HR Technology Vendors
The most immediate consequence is a new scrutiny requirement for AI screening tools offered as part of enterprise HR platforms. Vendors will need to demonstrate that their AI models can be audited for disparate impact across protected characteristics and that the decision logic is explainable to the standard required under employment discrimination law. Disparate impact analysis has been a requirement for employment selection tools under federal law for decades. The challenge with AI-based tools is that the training data, feature selection, and weighting logic are often proprietary and opaque, making it difficult for third parties or plaintiffs to demonstrate the specific mechanism through which bias occurs.
The Workday ruling opens the door to discovery that could force exactly that kind of transparency. A collective action that Workday itself has speculated could reach hundreds of millions of applicants creates significant incentive for plaintiffs to pursue full discovery of model training data, feature selection processes, and outcome statistics.
What This Means for the HR Technology Leader
For HR technology leaders, the ruling creates a due diligence requirement that did not exist with the same force before this decision. Organizations using AI-powered recruiting or screening tools, whether sourced from dedicated recruitment technology vendors or embedded in HCM platform suites, need to understand the disparate impact data for those tools and how the vendor handles bias audit requests.
The vendor questions are now operational. Buyers need to know whether the screening tool vendor maintains disparate impact statistics for its models, whether those statistics can be broken down by role type, geography, and candidate demographic, and what the vendor does when a customer identifies potential bias in screening outcomes. HR technology procurement conversations will need to cover all three.
For organizations that have deployed AI screening tools without a documented bias audit process, the Mobley ruling is a prompt to establish one before a complaint creates the urgency. Remediation after a complaint is filed is substantially more expensive and reputationally damaging than audit and mitigation conducted proactively. The California AI executive order, which directed agencies to recommend WARN Act revisions within 180 days to address AI-driven workforce impacts (as detailed in our earlier analysis of California’s compliance clock), establishes a parallel regulatory timeline that HR technology leaders should monitor alongside this legal development.
The AI hiring tool market has grown quickly on the promise of efficiency: faster screening, reduced recruiter workload, more consistent evaluation criteria. That promise is real. The Mobley ruling establishes that it does not override employment discrimination law, and that the vendor supplying the tool cannot disclaim responsibility for what the tool produces.
Source: U.S. District Court for the Northern District of California, Mobley v. Workday order
