Unveiling Algorithmic Bias: Workday Faces AI Hiring Discrimination Lawsuit

Table of Contents

• The Case 
• Legal Arguments
• Regulation in AI Hiring
• Conclusion
• References

As AI continues to advance and become more omnipresent in our everyday lives, we must understand that it, too, has racial and ethnic biases embedded in its software. Put simply, since AI hiring software learns from data from all corners of the internet—which for the most part has been written and annotated by human beings—it will also develop biases, conscious or unconscious, based on its input, similar to human learning. The question arises: how does this affect day-to-day life?

The Case

For finance graduate Derek Mobley, potential AI biases hindered his ability to secure a job. According to a Bloomberg Law article, Mobley, a 49-year-old Black man, claims that after applying for nearly 100 jobs through Workday Inc., an AI-driven human resources platform, he noticed a pattern of rejection, leading him to believe that the process was automated and discriminatory.

Specifically, in 2023, Mobley filed a class action lawsuit against Workday Inc., which alleges that its artificial intelligence systems discriminate against applicants who are Black, disabled, or over the age of 40. 

The article specifies that on January 9, 2024, the court case began with Workday attempting to dismiss it altogether, stating that their company commits to trustworthy AI hiring policies such as transparent design and delivery of AI solutions. The dismissal motion is still under review as of January 17, 2024. 

Legal Arguments

The case argument lies in algorithmic blackballing. The term ‘blackballing’ refers to the phenomenon of AI using historical data to learn from and generate, which can lead to discriminatory patterns—the phenomenon alleged in Mobley’s case. Specifically, Civil Rights Litigation Clearinghouse explained that the lawsuit is grounded in Title VII of the Civil Rights Act, 42 U.S.C. § 1981, ADA Amendments Act of 2008, and the Age Discrimination in Employment Act.

Derek Mobley seeks class certification, declaratory judgment, injunctive relief, back and front pay, interest, benefits, and damages, emphasizing the alleged discriminatory nature of Workday’s screening products.*

Mobley’s case is just the beginning of a complex national and global debate on using AI in the workplace. For example, according to Law360, the Equal Employment Opportunity Commission (EEOC) has released guidelines to prevent discrimination in automated hiring processes, settling its first automation-based case in August.

Regulation in AI Hiring

Additionally, New York City has enacted legislation to regulate AI in hiring, leading the way in addressing algorithmic bias or blackballing. 

The article also emphasized that similar cases have emerged in the healthcare sector, with the estates of Gene B. Lokken and Dale Henry Tetzloff suing UnitedHealth Group Inc. for using a flawed AI model to determine coverage criteria. Another lawsuit targeted State Farm, alleging discriminatory outcomes for Black homeowners due to biased algorithms in claims processing. 

Conclusion

As the legal landscape unfolds around AI, these cases pose significant questions about liability for biased AI products, shaping the future of AI-related legal precedents. This lawsuit and other suits shed light on the potential bias in AI hiring tools and raise critical questions about responsible AI practices.

Despite the outcome, the class action raises several ethical questions, as AI bias has now potentially penetrated the job market, perpetuating human-based recruiting biases rather than undoing them. One of the many purposes of AI is to improve on what we have already accomplished, including the work done to dismantle recruiting bias, not undo it. 

*Legal groundings for lawsuit summarized by AI from Civil Rights Litigation Clearinghouse