When AI Marketing Meets the Form 10-K

In May 2025, Workday announced new artificial intelligence agents designed to accelerate hiring, improve productivity, and transform human resources operations. The company emphasized that these technologies were built on its foundation of responsible AI.

But by the time of that announcement, Workday had already been disclosing litigation alleging discrimination involving its products and services for nearly two years.

The earliest specific disclosure I located appeared in Workday's May 25, 2023 Form 10-Q, approximately three months after the filing of Mobley v. Workday.

In that quarterly report, Workday acknowledged:

“We have been subject to a lawsuit alleging that our products and services enable discrimination”

The company disputed the allegations and warned that legal proceedings could be expensive, disruptive, and damaging to its reputation.

Significantly, the same disclosure discussed the possibility that customers' publicly reported analyses of AI-assisted employment decisions, including assessments required under New York City's automated employment decision tools law, could expose Workday to reputational or legal consequences.

The disclosure did not identify Mobley by name. Nevertheless, its timing and subject matter strongly suggest that it referred to that litigation.

Workday subsequently repeated and expanded the disclosure.

Its March 8, 2024 Form 10-K acknowledged the discrimination lawsuit and specifically addressed the possibility that litigation could extend to Workday customers.

Its March 6, 2026 Form 10-K again identified the litigation, now referring expressly to allegations that certain AI-related products and services enable discrimination.

This chronology is important: Workday did not wait until 2026 to acknowledge an existing discrimination lawsuit, nor did it describe discrimination exclusively as a hypothetical future risk.

Instead, the company continued expanding and promoting its AI offerings while consistently informing investors about an unresolved legal challenge.

What Workday Promises About Responsible AI

Workday's public representations extend beyond efficiency and productivity.

Its responsible AI materials describe human oversight, risk-based assessments, fairness testing, and monitoring intended to identify and reduce bias.

The company emphasizes that its technology assists rather than replaces human decision-making.

Workday also describes AI-generated candidate-match recommendations that recruiters can use to evaluate and prioritize applicants.

These distinctions are important.

An automated recommendation may substantially influence which applicants receive interviews or further consideration even when the software does not make the final hiring decision.

But the existence of such recommendations does not establish unlawful discrimination. That depends on the system's actual operation, its use by employers, and the applicable legal requirements.

Likewise, Workday's statements concerning responsible AI are representations about its development and governance practices. They do not independently establish whether every screening procedure satisfies federal employment discrimination laws.

The Unresolved Legal Questions

The Mobley litigation challenges the operation of automated applicant-screening technology, including allegations involving age and disability discrimination.

Workday disputes the allegations. No final judicial determination of unlawful discrimination has been made.

Nevertheless, the litigation raises questions about the relationship between Workday's publicly described AI governance practices and the operation of the technologies challenged in court.

What testing and validation supported the screening systems used during the relevant periods?

Did the evaluations examine characteristics that might indirectly reflect disability, age, or other protected traits?

How did the company assess the practical effects of candidate-ranking recommendations on employment opportunities?

Were relevant findings, complaints, and litigation developments incorporated into its continuing assessments of product performance and legal risk?

These questions do not presume wrongdoing.

They concern whether the company's representations about its AI capabilities and responsible development practices can be evaluated against evidence of how the relevant systems actually operate.

How Competitors Describe Similar Risks

Other enterprise-software companies also acknowledge potential AI-related legal exposure.

Oracle's 2026 Form 10-K identifies bias, discrimination, and lack of transparency among the possible risks associated with artificial intelligence.

ADP's 2026 annual report discusses discrimination, regulatory investigations, enforcement actions, and litigation arising from AI technologies.

Paycom's 2025 filing describes artificial intelligence as a competitive differentiator while recognizing evolving regulatory requirements and potential compliance burdens.

These companies offer different technologies, and their disclosures do not establish comparable liability or involvement in the Mobley litigation.

The comparison demonstrates that AI-related discrimination has become a recognized legal and corporate-reporting concern among major enterprise-software providers.

What Does the Disclosure History Establish?

Workday's filing history supports an important conclusion, but not necessarily the one suggested by an overly simple comparison of AI marketing and cautious risk disclosures.

The company acknowledged the discrimination litigation in May 2023, repeated that disclosure in subsequent filings, and continued promoting AI capabilities during the same period.

That history does not establish AI-washing, concealed litigation, or inadequate securities disclosures.

It does, however, provide a concrete basis for examining whether a company's promotional representations, technical evidence, operational practices, and assessment of material legal risks remain consistent over time.

For employers, technology vendors, directors, and shareholders, the central questions are straightforward.

What evidence supports representations that AI screening systems are reliable, appropriately validated, and designed to reduce discriminatory outcomes?

How are credible allegations and adverse findings investigated?

And as litigation develops, how does management determine whether changing risks require additional disclosures or financial recognition?

A company need not concede liability for disputed claims. Its public filings also need not reproduce every technical detail of its AI systems.

But material representations about AI capabilities should be supportable, and material legal risks must be evaluated under applicable accounting and securities-law requirements.

The issue is not whether a company's marketing materials and SEC filings emphasize different aspects of artificial intelligence. It is whether its representations about what the technology can do are supported by evidence—and whether its disclosures adequately address material risks arising from what the technology actually does.