www.loubar.org 16 Louisville Bar Briefs AI on Trial: Mobley v. Workday and the Future of Employment Law Marianna Michael Melendez and Catie A. Wheatley Artificial Intelligence (AI) has infiltrated almost every aspect of our lives. From being used as medical receptionists to conducting legal research, AI seems to know no bounds. And employment law is no exception. AI’s impact is evident from Mobley v. Workday, Inc., 3:23- CV-00770, an employment discrimination case currently pending in the Northern District of California. While Mobley is the bellwether case, this case is not just about AI or one lawsuit against one platform. There are potentially far-ranging implications regarding employment discrimination claims premised on novel theories of agency and disparate impact. Of course, AI has been criticized for its potential to reflect human bias from the very begin- ning, including its use in the employment sphere. For example, several large companies like Workday, Inc., the defendant in Mobley, use AI to screen job applicants. But it is well documented that AI platforms, such as ChatGPT, exhibit various biases when sorting resumes. The University of Washington conducted research to determine the extent of the discriminatory bias. See Stefan Milne, AI tools show biases in ranking job applicants’ names according to perceived race and gender, WASH. UNIV., Oct. 31, 2024, https://www. washington.edu/news/2024/10/31/ai-bias-resume-screening-race-gender/. The research discovered significant racial, gender and intersectional bias in how three state-of-the-art large language models ranked resumes. In conducting their research, the researchers used varied names associated with white and Black men and women across more than 550 real-world resumes and found the large language models favored white-associated names 85% of the time; male-associated names 52% of the time versus female-associated names only 11% of the time; and never favored Black male-associated names over white male-associated names. Id. Workday is now being accused of having these same inherent biases. Workday is a cloud-based financial management, human capital management and student information system software vendor. Workday launched in 2006 in California. Today, Workday is an international business that has clients in more than 175 countries. According to its website, 10,000 customers use Workday, from small businesses to more than 60% of Fortune 500 businesses. Workday offers a suite of products to help businesses manage their human resources among other things. And for job applicants, Workday stores resumes and other details to pre-populate future applications with the same employer to simplify the application process. Workday states that it “is a global leader and contributor of Responsible AI, committed to bringing to life Workday’s ethical AI principles and practices through our technology, our partner ecosystem, and broader global engagements.” Workday further states that each of its partners “on AI Marketplace must demonstrate adherence to specifications in the Workday Responsible AI Governance Framework, including the areas of ethical AI, data privacy & security, fairness, transparency, and quality.” Workday offers several AI apps and agents, which they call “Solutions,” that are run by Workday and various partners. For example, Workday offers Kainos Smart Test, Kainos Employee Document Management and Kainos Smart Audit. Kainos is a company that develops information technology for businesses and organizations, and each of these AI platforms boast of methods to streamline the hiring and employment process. To further its vision, Workday claims that the “workplace of the twenty-first century has been defined by rapid technological evolutions. From paper to computers to the cloud, forward- thinking businesses have always been at the forefront of that change. But AI technology isn’t just a future trajectory—it’s already an integral part of today’s business landscape. In order to stay ahead, it’s critical that businesses adopt AI technologies and the innovations they enable. For HR, AI has proven to be a decisive factor in enabling a streamlined recruiting process, personalized career development, and improved employee engagement—and that’s just the start.” Enter Mobley v. Workday, Inc., a cutting edge discrimination case, which challenges Work- day’s claims and points to the potentially dark underbelly of AI. Mobley, the plaintiff, first filed his lawsuit in February 2023. According to the first amended complaint, Mobley is an (1) African American male (2) over the age of 40 who holds a bachelor’s degree in finance from Morehouse College, an HBCU, and (3) experiences anxiety and depression. Mobley alleges that he applied to more than 100 positions since 2017 with various companies that use Workday’s applicant screening tools. According to Mobley, Workday’s algorithmic decision-making tools discriminate against applicants who are African American, over the age of 40 and disabled. Last year, Workday tried to dismiss Mobley’s amended complaint by arguing it was not a covered entity under the relevant statutes because it was not an employer making em- ployment decisions. The court dismissed Mobley’s claims that Workday was a covered entity because it was acting as an employment agency. But Mobley’s claims survived based on the court’s reasoning that the statutory language under Title VII, the ADA and the ADEA clearly extend to “any agent” of an employer and Workday was acting as an agent of the putative employers—a novel theory that was supported by a brief filed by the EEOC as amicus. The court also concluded that Mobley had sufficiently alleged a disparate impact claim, which does not require proof of intentional discrimination. According to the court, Mobley suf- ficiently alleged that he had disclosed his protected traits to Workday based on his allegation that Workday’s large language model could discern protected characteristics derived from demographic information based on zip code, college, membership in certain groups, year of graduation and work history. This disparate impact ruling is especially noteworthy given President Trump’s Executive Order issued earlier this year to eliminate EEOC enforcement based on disparate impact theory. In “Restoring Equality of Opportunity and Meritocracy,” President Trump claims that “a pernicious movement endangers this foundational principle, seeking to transform America’s promise of equal opportunity into a divisive pursuit of results preordained by irrelevant immutable characteristics, regardless of individual strengths, effort, or achieve- ment.” The Executive Order identifies “disparate-impact liability” as “a key tool of this move- ment.” However, this Executive Order neither overturns established case law nor impedes private litigation like Mobley’s. Additionally, this Executive Order does not stop state EEO agencies from pursuing such claims, which are well established under Title VII and other anti-discrimination statutes that piggyback off Title VII’s statutory language. Mobley v. Workday is still pending. In February, Mobley sought to expand his ADEA claim to a national class action, and the court granted preliminary certification this spring which allows Mobley, as lead plaintiff, to notify other alleged job seekers age 40 and above who applied through Workday’s service and were denied employment of their right to join the litigation. In the meantime, although Workday and similar applicant management vendors may be off the hook when it comes to being viewed as employment agencies, employers of any size should take note of the Mobley court’s reasoning regarding the statutory language under Title VII, the ADA and the ADEA that clearly extends to “any agent” of an employer. And the allegations in Mobley and early rulings are a reminder that employers face exposure from its vendors and AI tools that disproportionately harm protected groups, even absent malice or bias from the employer and its hiring managers. Bottom line—human oversight, documentation and transparency in the hiring process are more important than ever. As one of the first major court challenges to the use of algorithmic hiring tools under federal employment discrimination laws, this lawsuit is one to watch. Marianna Michael Melendez is the Chair of the LBA Labor and Employment Law Section. Marianna is a University of Louisville Brandeis School of Law graduate and has been practicing labor and employment law for the last six years. She is currently the Jefferson County Public School’s Assistant General Counsel for Compliance and Investigations. Catie A. Wheatley is the Vice-Chair of the LBA Labor and Employment Law Section. Catie is a litigation attorney at Faegre Drinker Biddle & Reath LLP who helps regional, national and global organizations assess risk and resolve business disputes whether arising from restrictive covenants or other employee related claims. In her free time, Catie enjoys reading, playing tennis and exploring the national parks with her spouse and child. n PROFESSIONAL EXCELLENCE Stop Guessing. Start Growing. G R E E N CA R DI G AN MAR K E T I N G Office in Louisville Law Firm Marketing Done Right. Your law firm’s marketing is WEAK. We clean it up with a BOLD strategy that delivers revenue. Find Out How We Help
