If your organization uses AI or automated tools to screen, rank, or evaluate job applicants, recent developments from the Northern District of California and the Connecticut legislature deserve your attention. Out of California, Mobley v. Workday, Inc, No. 23-CV-00770-RFL, is now in its third year of litigation and continues to produce significant rulings, grappling with the question of how existing federal anti-discrimination laws apply to AI driven hiring systems. In Connecticut, the state legislature’s recent passage of SB 5, Connecticut Artificial Intelligence Responsibility and Transparency Act, reflects a growing trend toward increased regulation of the use of automated employment hiring tools.  Together, these developments confirm that the legal risks associated with AI in hiring are real, growing, and no longer theoretical.

Mobley v. Workday at a Glance

At its core, the case poses a straightforward question that should concern every employer using automated hiring tools: who is liable when an AI-driven screening system produces discriminatory outcomes?

Workday, Inc. provides a platform used by thousands of employers to process job applications, often serving as the gateway through which applicants must pass to even be considered for a role. The plaintiffs allege that the platform’s AI-driven screening and ranking tools are producing discriminatory outcomes.

Here are the key facts:

• The case was originally filed in 2023 by Derek Mobley, an African American applicant over 40, who alleges he was repeatedly rejected for jobs through Workday’s platform. The complaint has since expanded to include additional plaintiffs spanning multiple protected classes.
  
• The claims invoke Title VII of the Civil Rights Act, the Age Discrimination in Employment Act (“ADEA”), and the Americans with Disabilities Act Amendments Act (“ADAAA”), along with numerous state law claims. The underlying theory is algorithmic bias, i.e., that AI systems can produce discriminatory outcomes even without anyone intending to discriminate.
  
• According to the complaint, Workday’s AI systems are trained on historical hiring data that may replicate patterns embedded in past decisions. Even where protected characteristics are not directly used as inputs, the system may rely on proxy data: employment gaps, years of experience, educational background, and career trajectories can all correlate with characteristics like disability, age, race, or gender. The complaint further alleges that Workday’s 2024 integration of HiredScore’s AI recruitment tools further embeds automated scoring into the decision-making process. 

• The plaintiffs also challenge personality and assessment tools integrated into the platform, arguing that they are not neutral screening mechanisms but data-gathering instruments that feed into broader candidate profiles. In some instances, the complaint alleges that these assessments may function like prohibited medical inquiries under the ADA, because they can infer disabilities through proxy indicators like illness-related absences and health-associated behavioral patterns.

Where the Case Stands: The Disparate Impact Theory Survives

The case is now over three years old, and the court has issued several significant rulings. In 2024, the court dismissed the intentional discrimination claims but allowed the disparate impact claims to proceed—a distinction with major practical implications (more on that below). The complaint has since been amended multiple times. Most recently, plaintiffs filed their Third Amended Complaint on March 27, 2026, adding new named plaintiffs, expanding the race claims to include Asian Americans, broadening the disability class to include physical disabilities, and strengthening the California Fair Employment and Housing Act (“FEHA”) claims with specific allegations that Workday’s screening systems were designed, trained, and controlled from its California headquarters. Workday filed its latest  motion to dismiss on April 10, 2026, arguing that plaintiffs exceeded the court’s limited leave to amend, that certain claims are time-barred, and that the FEHA allegations remain insufficient. The motion is now fully briefed and awaiting decision.

Regardless of how that motion is resolved, the key takeaway for employers is this: the core federal disparate impact claims have survived every challenge to date. Workday has also sought certification of an interlocutory appeal on the question of whether job applicants can bring disparate impact claims under the ADEA—a sign that the legal questions in this case are being contested at the highest levels and remain unsettled.

This distinction matters. Disparate impact claims focus on outcomes, not intent. If your hiring process disproportionately screens out members of protected groups, liability can arise even without any proof of discriminatory motive—unless you can show that the practice is job-related and consistent with business necessity.

The plaintiffs are also pointing to audit data that allegedly shows statistically significant disparities affecting certain groups. If that evidence holds up, it could play a major role as the case advances, particularly in connection with class and collective certification.

Why This Matters for Your Connecticut Organization

Even though Mobley is pending in California, the legal issues extend nationwide. For Connecticut employers in particular, there is an additional layer of risk. Connecticut’s Fair Employment Practices Act (“CFEPA”) provides broad protection against employment discrimination, and in some cases, plaintiffs need only show that discrimination was a “motivating factor,” a lower bar than what federal law requires. 

On top of that, on May 1, 2026, the Connecticut General Assembly passed Senate Bill 5, the Connecticut Artificial Intelligence Responsibility and Transparency Act. Among other things, SB 5 will require employers using AI tools that materially influence employment decisions to provide applicants and employees with plain-language disclosures and pre-decision notices before those tools are used to make decisions about them. The bill also amends the CFEPA to state explicitly that relying on an automated tool is not a defense to a discrimination claim. We published a detailed breakdown of SB 5’s employer-facing provisions, effective dates, and compliance steps in a separate post, which can be found here.

Put simply: if you use an AI hiring tool and it produces discriminatory results, your organization may still be liable—even if the tool was developed by a reputable vendor, even if you did not intend to discriminate, and even if you had no visibility into how the tool works.

What Connecticut Employers Should Do Now

The takeaway from Mobley and from Connecticut’s new AI legislation is not to abandon AI in hiring—these tools offer real efficiencies—but to use them with appropriate oversight. Here are concrete steps to consider:

• Know your tools. If AI is involved in screening or ranking candidates, you should know what data is being used and how decisions are being made. Black-box systems, where the logic behind a decision is opaque even to the employer, create significant exposure.
• Review your vendor relationships.  The Mobley case illustrates how rapidly AI hiring platforms evolve, as Workday’s 2024 integration of HiredScore added another layer of AI-driven processing to be scrutinized. Review contracts to understand what representations the vendor is making about the tool’s compliance, what testing has been conducted, and where responsibility falls if problems surface. Indemnification provisions, audit rights, and data-sharing obligations all merit close attention.
• Audit your assessments. If your hiring process includes assessments or personality tests, examine whether any of them could be characterized as disability-related inquiries under the ADA. The line between a permissible pre-employment screening tool and an impermissible medical inquiry is one the courts are actively defining.
• Prepare to justify your processes. If a tool produces a disparate impact, you will need to demonstrate that it is job-related and consistent with business necessity. That analysis should happen proactively, not for the first time after a claim has been filed.
• Document your diligence. Keep records of the steps you have taken to evaluate and monitor your AI hiring tools, including any vendor correspondence, audit results, and bias testing. If your processes are ever challenged, that documentation will be your best evidence of good faith.
• Prepare for SB 5 compliance. If your organization uses AI tools in hiring or other employment decisions, begin mapping those tools now. Identify which systems produce outputs that materially influence decisions, review your vendor contracts, and start building disclosure and notice workflows. Our separate post on SB 5 provides a detailed compliance roadmap.

Mobley v. Workday is one of the first major cases to test whether existing anti-discrimination law can hold AI hiring tools accountable. Connecticut’s SB 5 represents the legislative side of that same trend. Employers who take the time now to understand their tools, audit for bias, and build compliance workflows will be in a significantly stronger position when—not if—these issues arrive at their door.

If you have questions about how Mobley v. Workday or how SB 5 may affect your organization’s hiring practices, or if you would like assistance auditing your AI hiring tools for compliance, please contact a member of Shipman & Goodwin’s Employer Defense & Labor Relations practice group. You can also listen to our recent From Lawyer to Employer podcast episode on AI in the workplace.