A federal judge in San Francisco has dealt a significant setback to Workday, the California-based human resources software giant, by allowing a landmark discrimination lawsuit to move forward. U.S. District Judge Rita Lin rejected Workday's arguments that it should not face liability for allegedly biased hiring practices, ruling that the company cannot escape accountability by claiming its screening decisions affected people applying for jobs outside California. The decision, handed down on Monday, represents a critical moment in technology accountability and employment law, as it opens the door for the first major class action lawsuit directly challenging the algorithmic decision-making embedded in AI recruitment tools.
The lawsuit, filed in 2023, takes direct aim at how Workday's software screens job candidates by filtering out applicants in ways that allegedly violate both California's anti-discrimination statutes and the federal Americans with Disabilities Act. Judge Lin's ruling on Monday dealt primarily with Workday's motion to dismiss amended claims in the case, and the judge largely sided with the plaintiffs. Crucially, Lin determined that because Workday operates from California and allegedly orchestrated the unlawful screening practices from its headquarters, the company can be held responsible under state discrimination law regardless of where applicants were located or where the jobs were situated. This geographical argument had been central to Workday's defense strategy.
The scope of the allegations is broad and reflects growing anxieties about artificial intelligence's role in employment decisions. Plaintiffs contend that Workday's software has systematically weeded out candidates based on protected characteristics, including disability status. More specifically, the lawsuit argues that the algorithm uses "proxy indicators"—indirect signals that correlate with disabilities or chronic illness—to filter candidates. Employment gaps, for instance, might suggest a health-related absence, and the system may penalise such gaps without any explicit disability filter. Judge Lin refused to dismiss this claim, finding it legally viable. Beyond disabilities, the plaintiffs separately allege discrimination against Black job seekers, women, and workers over 40 years old, all groups protected under federal employment law.
One claim did not survive Monday's ruling: allegations that Workday's software discriminated specifically against Asian American applicants. Judge Lin dismissed this claim on procedural grounds, finding that the plaintiffs had not followed proper legal steps to add it to the lawsuit, though this decision leaves the door open for potential resubmission if proper procedures are followed. The ruling on the Asian American claim, however, is narrow and does not represent a finding on the merits of such allegations.
This case arrives at a moment when AI-driven hiring has become ubiquitous across American business. Industry surveys consistently show that more than 80 per cent of United States employers now deploy artificial intelligence tools in recruitment, and virtually every Fortune 500 company relies on such technology. Workday's software is among the most widely used, making this litigation potentially consequential for how thousands of organisations conduct hiring. The proliferation of these tools reflects corporate appetite for efficiency and consistency, yet legal scholars and worker advocates have repeatedly warned that automation can perpetuate and amplify existing biases embedded in training data.
Despite widespread adoption, litigation over algorithmic hiring has remained sparse until now. Experts attribute this gap partly to information asymmetry: most job applicants never discover whether employers used AI to screen them out, making it difficult to recognise when discrimination might have occurred. Additionally, suing over novel technology involves complex expert testimony and unfamiliar legal territory, creating barriers to litigation that traditional employment discrimination cases do not face. The Workday case, therefore, potentially signals a shift toward greater legal scrutiny of hiring algorithms.
Government agencies and worker advocates have long flagged concerns about AI recruitment tools. The worry is not merely theoretical: when algorithms are trained on historical hiring data, they can absorb patterns of past discrimination. If an organisation previously hired fewer women or older workers in certain roles, the algorithm learns those patterns and replicates them, all while appearing objective and data-driven. This mechanism makes algorithmic discrimination particularly insidious because it operates without human judgment and can therefore escape detection longer than explicit bias would.
Judge Lin's earlier decision to reject Workday's initial motion to dismiss, combined with Monday's ruling on amended claims, suggests the court believes the plaintiffs have stated plausible allegations worthy of proceeding to discovery and potential trial. The judge's reasoning—that Workday's California operations made it subject to California law even when screening remote applicants—also has implications beyond this single case. It suggests that companies cannot geographically arbitrage liability by conducting unlawful conduct from one state while claiming they lack connection to harmed applicants elsewhere.
For Malaysian and Southeast Asian readers, this case holds significance beyond United States borders. Many multinational corporations operating in Malaysia and the region use the same or similar AI hiring platforms that Workday supplies. If the California lawsuit succeeds and establishes that such tools can violate anti-discrimination law, it will likely influence how regional subsidiaries and regional offices of global companies conduct recruitment. Moreover, as Southeast Asian economies increasingly adopt advanced technologies, local regulators may draw lessons from this litigation about how to govern algorithmic employment decisions before problems become widespread.
The path ahead for this litigation remains lengthy. Discovery will require detailed examination of Workday's algorithms, training data, and outcomes. Experts will need to analyse whether the software's decision-making patterns correlate with protected characteristics in statistically significant ways. Workday will argue that any disparate impact results from legitimate, non-discriminatory factors, not bias in the algorithm itself. The company and its legal representatives have not yet publicly commented on Monday's ruling.
This case ultimately represents a reckoning that was probably inevitable: as artificial intelligence becomes central to consequential decisions affecting people's livelihoods, the legal system must grapple with how to hold creators accountable when those systems cause harm. Whether Workday ultimately prevails or loses, the lawsuit will force disclosure of algorithmic decision-making that has largely operated in opacity, setting precedent for how employment discrimination law adapts to the era of automation.
