As artificial intelligence (AI) has become more deeply integrated into the operations and investor communications of public companies, it has generated a corresponding wave of securities litigation. Much of the litigation has centered on AI-washing allegations or overstating the extent to which a company was utilizing AI. As we discussed in our article from last November, More Buzz Than Sting, early data suggested that these cases largely followed familiar patterns. Although filing activity rose modestly in raw numbers, dismissal rates, settlement values and legal theories remained broadly consistent with the wider securities litigation landscape.
That pattern, however, has begun to evolve. Plaintiffs no longer appear to be predominantly focused on AI-washing. Instead, as companies incorporate the technology into strategic planning, capital allocation, product development and operational execution, allegations are now arising from those activities. For directors and officers, as well as risk and legal teams, their attention must turn now to making informed strategic decisions and managing the disclosures and investor expectations that accompany AI-driven initiatives.
AI as context rather than cause
Consistent with the observations in our November publication, early AI-related securities class actions followed a predictable pattern: misleading statements about the extent that AI technology was being used, a corrective disclosure and a corresponding stock price decline.
Recent filings, however, reflect a broader landscape, where claims have arisen from more nuanced disputes over guidance, investment decisions, operational performance and business execution in which AI forms part of the underlying business strategy. This observation is not new. Among others, Kevin LaCroix and Sarah Abrams have noted across several D&O Diary posts (see, as examples, here, here and here) a distinction between two types of AI-related securities litigation: (1) cases alleging that a company overstated its AI capabilities, and (2) the emergence of newer cases alleging it understated its AI-related risks.
Several recent cases illustrate this change.
- Telus International: Perhaps the earliest case marking this evolution was Telus International in January 2025. Investors there alleged that the digital services company misled the market by overstating the maturity and profitability of its AI offerings and by failing to disclose that its AI strategy was undermining its core business. The complaint claimed, in part, that AI initiatives consisted of low-margin pilot projects that diverted resources from higher-margin legacy services, increasing costs and compressing margins. Management purportedly knew that AI efforts were cannibalizing established revenue streams and eroding profitability yet continued to present the technology as a growth and efficiency driver.
- Upstart Holdings: In April 2026, plaintiffs filed a securities class action against Upstart Holdings and certain of its officers and directors. They alleged the company's AI-driven lending model failed to perform as expected under changing macroeconomic conditions, ultimately contributing to reduced loan volumes and a downward revision in revenue guidance. Rather than alleging the company overstated its AI capabilities, plaintiffs contended that management failed to adequately disclose risks arising from the company's use of AI within its lending model and the potential impact of those risks on the business and its financial outlook.
- Oracle Corporation: Oracle became the subject of a securities class action in March 2026. In that case, plaintiffs alleged that the company made misleading statements and omissions concerning its AI infrastructure strategy, including the expected economic benefits of substantial investments in AI-related data-center capacity. Plaintiffs claimed that Oracle failed to adequately disclose that the strategy would require significantly increased capital expenditures and financing commitments without corresponding near-term revenue growth. This purportedly created risks to free cash flow, debt levels and project funding. Litigation followed after concerns emerged regarding the costs and economics of the company's AI infrastructure buildout.
- Commvault Systems: A similar dynamic appeared in the May 2026 securities class action involving Commvault Systems. The company emphasized its AI-enabled cyber-resilience platform and related growth opportunities while shifting toward SaaS and AI-driven offerings. According to the complaint, that transition altered the economics of the company's sales mix, as SaaS transactions carried lower average selling prices and affected annual recurring revenue calculations. Plaintiffs asserted that these changes contributed to the company's failure to meet projected net new ARR growth targets and that the impact of the shift was not adequately reflected in the company's financial guidance and disclosures.
- Power Solutions International: The March 2026 case involving Power Solutions International arose from a similar set of issues. There the company pivoted toward supplying power systems for data centers benefiting from AI-driven infrastructure demand and attributed growth and margin expansion to that opportunity. It later disclosed declining margins resulting from manufacturing and supply-chain inefficiencies associated with rapidly scaling production for data-center product lines. The resulting stock price decline prompted a securities class action alleging that the company overstated its ability to capture data-center demand and understated the operational costs and challenges associated with expanding manufacturing capacity.
From buzz to embedded risk
The characterization in our November publication of early AI-related litigation as “more buzz than sting” reflected the view that, despite considerable attention to AI as a corporate risk, the available evidence did not suggest a meaningful departure from customary securities litigation patterns.
The more significant development is AI's growing influence on the business decisions that shape corporate performance. As AI becomes embedded in business activity, the consequences of those decisions increasingly become the subject of broader, more traditional disclosure litigation.
The through line for the relationship between AI and litigation risk may be this: AI is a disrupter, forcing companies to adapt and change the way that they do business. Changes to business models need to be accurately disclosed; in fact, the failure to make changes to a business model may also need to be disclosed. At the moment, stakeholders are particularly interested in and seeking disclosures that relate to AI. When companies are forced to make them, there are increased opportunities for footfalls or worse.
To the degree that organizations weave AI into strategy and operations, the outcomes it drives, rather than the technology itself, are increasingly what give rise to securities claims. If a meaningful “sting” emerges, it will come from those outcomes, not from any shift in underlying liability theories. The implication is that AI risk is an exposure that must be addressed through the same governance, disclosure discipline and execution rigor that support all material business decisions.
Disclaimer
WTW hopes you found the general information provided here informative and helpful. The information contained herein is not intended to constitute legal or other professional advice and should not be relied upon in lieu of consultation with your own legal advisors. In the event you would like more information regarding your insurance coverage, please do not hesitate to reach out to us. In North America, WTW offers insurance products through licensed entities, including Willis Towers Watson Northeast, Inc. (in the United States) and Willis Canada Inc. (in Canada).
Authors
