Responsible AI and Bias Mitigation for HiredScore Spotlight
At Workday, we understand that responsible AI requires rigorous evaluation, independent review, and transparency about how our tools perform in practice. As part of that commitment, we asked Secretariat, an independent third party, to conduct an analysis of Workday’s deployment of HiredScore Spotlight using Workday’s own applicant flow data.
We are publishing a summary of that analysis to provide greater transparency into Workday’s responsible AI practices and to support customer trust in Workday as a responsible developer and deployer of AI tools.
This page is provided for transparency purposes only. It is not intended to satisfy any customer-specific legal or regulatory obligation, nor does it suggest or imply that HiredScore Spotlight is or is not an automated employment decision tool under any particular law.
Independent analysis of Workday’s deployment of HiredScore Spotlight
Secretariat conducted the analysis at Workday’s request using data collected in connection with Workday’s deployment of HiredScore Spotlight. The analysis was designed to assess outcomes across demographic groups using comparative measures commonly used to evaluate potential differences in rates across categories.
The table below summarizes the results of that analysis. It includes candidate volume and outcome-rate information across relevant demographic categories, as well as comparative ratios used to identify whether differences may warrant further review.
Because these results are based on Workday’s own deployment context and data, they should be understood as specific to Workday’s implementation of Spotlight during the relevant time period. Outcomes may differ in other settings depending on job requirements, candidate pools, workflow design, configuration choices, and human review processes.
Why we are sharing these results
We are sharing this information because transparency is an important part of responsible AI. Customers and other stakeholders should be able to understand not only that Workday has governance processes in place, but also that we subject our tools to independent analysis and share meaningful information about the results.
Publishing this summary is one way we demonstrate our commitment to building and deploying AI tools responsibly and to earning trust through evaluation, documentation, and openness.
Analytic Results Summary
Date of Testing Completion: March 20, 2026
Distribution Date: September 1, 2025 - February 28, 2026
Data Collected Description: Spotlight simplified output showing match grades for candidates reported qualifications against Workday’s specific job requirements, in an A, B, C, D grading system where A indicates the closest match. The analysis was conducted on a sample of Workday's own applicant data from the most recent six (6) months of Spotlight deployment at the time of analysis.
The analytic dataset includes applications from U.S. residents for positions in the greater New York City area submitted between September 1, 2025 and February 28, 2026. For this analysis, we only included the five job profiles with the highest volume of applications. Furthermore, we only included job profiles where there were enough applicants in each group to support a meaningful comparison. We excluded applicants who did not provide gender or race information or who provided conflicting information across multiple applications.
Source of Data Collected: The Applicant Tracking System (ATS), which includes Spotlight grades for applicants.
Audit Outcome: There is no evidence of disparate impact based on the calculated impact ratios presented below.
