By Gerald L. Maatman, Jr., Adam D. Brown, and Elizabeth G. Underwood
Duane Morris Takeaways: In Mobley, et al. v. Workday, Inc., Case No. 23-CV-00770, 2026 WL 1510537 (N.D. Cal. May 29, 2026) (ECF No. 340), Magistrate Judge Laurel Beeler of the U.S. District Court for the Northern District of California issued an order resolving three discovery disputes in this closely watched employment discrimination class action involving novel artificial intelligence (AI) issues. The Court denied Plaintiffs’ motion to compel production of Workday’s bias-testing data, finding that the attorney-client privilege protects the data because Workday’s attorneys curated it and used the results in providing legal advice. The Court also denied Plaintiffs’ motion to compel Workday to produce its customers’ applicant data because Plaintiffs failed to show that Workday had control of that data within the meaning of Rule 34 of the Federal Rules of Civil Procedure. However, the Court ordered production of Workday’s EEO-1 and Office of Federal Contract Compliance Programs (OFCCP) documents, finding those documents to be relevant to Workday’s knowledge of potential demographic disparities when utilizing its AI tools.
The ruling is significant for corporate counsel. For employers navigating the intersection of privilege, discovery obligations, and AI hiring tools, this ruling provides important guidance on protecting bias-testing data while recognizing the broad scope of discoverable information in AI employment discrimination cases.
This development follows Workday’s unsuccessful Motion to Dismiss Plaintiff’s Amended Complaint, which we blogged about here, Workday’s first successful Motion to Dismiss, which we blogged on here, and the EEOC’s amicus brief filing, which we blogged about here.
Case Background
Plaintiffs are suing Workday for utilizing an AI screening system that allegedly is more likely to deny employment applications from individuals who are African American, suffer from disabilities, or are over forty years old. Id. at *1. Workday Recruiting is a software product that helps customers manage hiring, and customers who purchase Workday Recruiting have access to an algorithmic feature called Candidate Skills Match, which determines the extent to which an applicant’s skills match the role to which they applied. Id. In 2024, Workday acquired HiredScore, which allowed Workday to offer additional features to customers, including Spotlight, a candidate review tool, and Fetch, a sourcing tool that connects organizations with potential talent by suggesting individuals for open jobs. Id.
As to the present discovery disputes, first, Plaintiffs filed a motion to compel Workday to produce its bias-testing data and its customers’ applicant data. Id. at *3. The parties disagreed as to whether the bias-testing data was protected by attorney-client privilege and whether Workday had control of its customers’ applicant data. Id. Second, Plaintiffs sought to compel production of Workday’s EEO-1 and OFCCP documents, with the parties disputing relevance, burden, and waiver. Id. at *6. Third, Plaintiffs moved to compel Workday to provide deanonymized data of applicants’ names and other application information. Id. at *7.
The Court’s Decision
Attorney-Client Privilege Applied To Bias-Testing Data
First, the Court agreed with Workday that its bias-testing data was protected from disclosure by the attorney-client privilege. Id. at *4. Specifically, the Court reasoned that the bias-testing data was privileged because Workday had shown more than mere direction from its attorneys and “ha[d] represented that its attorneys curated the data it used in the bias testing, the overall purpose of the testing was to provide legal advice and not to be used in a business capacity, and it ha[d] not submitted the data to a regulatory body.” Id.
Moreover, the Court rejected Plaintiffs’ arguments that Workday had waived privilege by using the bias-testing data offensively through reliance on an “AI Fact Sheet” that stated Workday performs bias testing. Id. at *5. Instead, the Court held that “Workday’s invoking the mere existence of its bias testing outside of litigation [was] not enough to waive privilege.” Id.
No Control Over Customer Application Data
Second, the Court denied Plaintiffs’ motion to compel Workday to produce its customers’ applicant data. Id. at *6. The Court found that Plaintiffs had not met their burden of demonstrating that the provision of the Master Subscription Agreement allowing Workday to produce a customer’s data under a court order constituted “control” under Rule 34 because Workday did not have a legal right to obtain its customers’ data on demand. Id. at *6. However, the Court observed that some third parties that Plaintiffs had subpoenaed had taken the position that Plaintiffs should seek the data from Workday instead. Id. Thus, the Court encouraged the parties to work together to resolve the issue. Id.
Production Of EEO-1 and OFCCP Documents
Third, the Court ordered production of Workday’s EEO-1 and OFCCP documents, finding that Plaintiffs had met their initial burden on relevance. Id. In particular, the Court reasoned that Workday utilizes the same AI tools as its customers, and under either the agent or direct-employer theory, “Workday’s EEO-1 and OFCCP documents are relevant to its knowledge of potential demographic disparities when utilizing AI tools.” Id. at *6.
Deanonymized Applicant Data
Finally, the Court disposed of Plaintiffs’ request for deanonymized applicant data as moot because Plaintiffs had admitted in subpoenas seeking the same information from third parties that they did not need applicant names. Id. at *7.
Implications For Employers
This decision reinforces the concept that bias-testing data can be shielded from production under attorney-client privilege when an employer’s attorneys curate the underlying data and conduct bias-testing for the purpose of providing legal advice, as opposed to a business or regulatory compliance purpose. Of note, and as supported by this Court’s decision, companies that utilize AI in their hiring processes should structure their bias-testing under the direction of legal counsel to preserve attorney-client privilege.
Moreover, the Court’s ruling on EEO-1 and OFCCP documents suggests that employers and AI vendors should be aware that they may face broad discovery obligations regarding their own use of the same AI tools they market to customers, as in this case, the Court found Workday’s EEO-1 and OFCCP documents relevant because Workday uses the same AI tools as its customers.
