Securities and Derivative Litigation: Quarterly Update

Artificial Intelligence-Related Securities Class Actions on the Rise

Background

Artificial intelligence (“AI”) remains a preeminent topic in the technology and business sectors, and more recently in courts of law as it has become the basis of numerous U.S. securities fraud class actions. In 2024, increased spending on AI-related research and development has been accompanied by a heightened regulatory focus on AI-related risks and disclosures by the SEC,¹  as well as an increase in litigation initiated by shareholder plaintiffs.²  In 2024, there were thirteen securities class action lawsuits filed in U.S. District Courts; more than double the amount filed in 2023.³  Despite the relative novelty of AI-based securities class action claims, we expect this trend to continue.

Jurisdiction of AI-Related Securities Class Action Cases

With the exception of January, at least one AI-related securities class action case has been filed each month this year: February (2); March (1); April (1); May (1); June (1); July (2); August (2); September (1); and October (2). Of the thirteen cases, four have been filed in the Northern District of California, home to many companies at the forefront of AI technology.⁴  Another five cases are pending in the Southern District of New York.⁵ And the remaining four are being litigated in the District of New Jersey; the District of Massachusetts; the Southern District of Texas; and the District of Maryland.⁶

Types of AI Defendants and Allegations

AI-related securities fraud class actions involve a variety of defendants, including both AI-focused companies and companies that use AI as one of many tools to further their business interests. The primary allegations raised in each of the complaints is that the defendants violated §§ 10(b) and 20(a) of the Exchange Act of 1934 and Rule 10b-5 promulgated thereunder by making materially false or misleading statements or omissions, with the requisite scienter, which caused investors to suffer financial losses. The majority of cases from 2024 are still in the early stages of litigation.

Several of the AI-related securities fraud suits allege that defendants overstated their AI capabilities or misrepresented the amount of future revenue to be attributed to AI-powered products or services. For example, in June of 2024, plaintiffs filed a securities fraud suit against ODDITY Tech Ltd., which plaintiffs describe as a beauty and wellness tech-platform that “purports to serve customers worldwide through its artificial intelligence-driven platform, using data science, machine learning, and computer vision capabilities to identify consumer needs [and] develop solutions in the form of beauty and wellness products.”⁷ The suit arose from a short seller report issued by the Ningi Research group, which indicated that despite referencing “AI” in its F-1 prospectus forty times, ODDITY used product-matching technology that was akin to normal, non-AI powered questionnaires, and the company’s references to repeat purchases based on its AI capabilities were actually attributed to performance marketing akin to that used by its competitors as well as non-cancelable plans.⁸ Following publication of the Ningi report, the complaint alleges that ODDITY’s share price fell, and shareholders filed suit.⁹

Other AI-related securities fraud complaints alleged that the companies made material misrepresentations related to their prospective growth and future sales based on the use of AI-powered products. For example, in June 2024, shareholders filed suit against UiPath, Inc., a company that provides business automation software powered by Robotic Process Automation (RPA) and AI.¹⁰ The complaint alleges that after experiencing plateaued growth and decreases in demand for its RPA products, the company announced a “turnaround” strategy in September of 2022.¹¹ The strategy consisted of rebranding UiPath as an AI-Powered Business Automation Platform and revising its sales strategy.¹² Shareholders contend that in describing the success of the turnaround strategy, defendants represented that UiPath could close more lucrative deals and that its AI-powered products set the company apart from the competition.¹³ UiPath purportedly experienced a strong third quarter in 2024, which the company and its representatives attributed to the turnaround strategy via a press release.¹⁴ According to the complaint, as a result, UiPath stock increased from US$19.76 to US$25.04 overnight.¹⁵ The company also announced revenue guidance of US$1.5B, attributing its success to the turnaround strategy.¹⁶ Plaintiff alleges that on May 29, 2024, the CEO of UiPath made an unexpected departure and UiPath announced disappointing financial results for the first quarter, cutting its revenue guidance by 10%.¹⁷ Defendants attributed the downturn to the company’s inability to execute a strategy to scale AI-powered growth products and confusion that customers were experiencing surrounding AI.¹⁸ The stock purportedly declined to US$6.23 per share and shareholders filed suit claiming, in part, that defendants misrepresented the success of the turnaround strategy.¹⁹

Alongside private shareholder suits, the SEC has also remained vigilant in prosecuting investment advisers charged with making false and misleading statements related to the use of AI.²⁰

Key Takeaways

The complaints filed by shareholder-plaintiffs in AI-related securities class actions provide helpful insight for entities engaging in the use and marketing of AI capabilities. First, while AI itself is a novel tool, the current enforcement framework for alleged securities fraud remains a vehicle for litigating issues arising from false or misleading statements or omissions, including those related to the use of AI. Second, statements touting a company’s unparalleled AI capabilities and the anticipated positive impact on a company’s growth should be made with caution as such statements are susceptible to litigation if things do not go as planned. And third, the use of risk disclosures and tempered expectations in public statements might serve as a method of properly accounting for risk related to the use of AI.

While securities class actions related to AI have continued to increase, it is important to note that many of these cases remain in the early stages of litigation. It will be interesting to examine how these claims are resolved and how regulatory agencies address concerns surrounding AI, which continues to represent both risk and opportunity.

Conclusion

While securities class actions related to AI have continued to increase, it is important to note that companies remain in the early stages of litigation. It will be interesting to examine how these claims are resolved and how regulatory agencies evolve to address concerns surrounding AI, which continues to represent both risk and opportunity.

Court Denies Class Certification in Case Brought Against Rocket Companies, Inc. in a Post-Goldman Sachs Ruling

On September 30, 2024, U.S. District Judge Thomas L. Ludington of the United States District Court for the Eastern District of Michigan relied on Goldman Sachs Grp., Inc. v. Arkansas Tchr. Ret. Sys.,²¹ in denying a renewed motion for class certification brought by investors of Rocket Companies, Inc. (“Rocket”), a constellation of 13 separate companies, spanning across multiple industries, including home and personal financing, sales, technology, and marketing.²²

Earlier in the year, Plaintiffs Carl Shupe and Construction Laborers Pension Trust for Southern California (“SoCal”) filed an Amended Complaint on behalf of themselves and all other similarly situated Rocket shareholders. The complaint included allegations that Rocket Holdings, Inc., which controlled Rocket, violated the Exchange Act and Rule 10b-5 when Defendant Dan Gilbert—RHI’s majority shareholder and Chairman—sold shares of Rocket Class A common stock armed with material, non-public information about Rocket’s financial forecast. Plaintiffs also alleged that Defendants RHI and Gilbert effectively controlled Rocket and are thus liable under Section 20(a) of the Exchange Act.²³

In Judge Ludington’s opinion, he denied Plaintiffs’ motion for class certification for a myriad of reasons, including that under the Supreme Court’s 2021 ruling in Goldman Sachs, Defendants rebutted the Basic fraud-on-the-market presumption of reliance.²⁴  In setting forth his analysis, Judge Ludington pointed to the decade-long class certification saga of Goldman Sachs, which reinforced that where plaintiffs have adequately invoked the fraud-on-the-market presumption of reliance, defendants may rebut the presumption by establishing, by a preponderance-of-the-evidence, a lack of price impact. Specifically, the court clarified how the nature of the alleged misrepresentations affects the price impact analysis. Indeed, the Court in Goldman Sachs had concluded that “the generic nature of an alleged misrepresentation often will be important evidence of price impact because, as a rule of thumb, a more-general statement will affect a security’s price less than a more-specific statement on the same question.”²⁵

Judge Ludington saw similarities between the Goldman Sachs case and the instant Rocket case. The Court found that in Rocket, like in the Goldman Sachs case:

• Plaintiffs alleged the CEO’s statements that Rocket would be financially successful were “front-end misrepresentations” that “artificially inflated or maintained” the price of Rocket Class A common stock;

• Plaintiffs presented indirect evidence—through conclusory allegations—that Rocket’s May 5, 2021 revelation about declining margins and closed-loan volume caused a 17% “back-end price drop” in Rocket stock;

• A considerable mismatch existed between the generic nature of the alleged misrepresentations and the specific revelation; and

• Defendants produced expert testimony suggesting that no analyst referenced Defendant Farner’s alleged misrepresentations in reports issued throughout the Class Period, which “sever[s] the link between” Plaintiffs’ alleged “back-end price drop” and their alleged “front-end misrepresentation[s].”²⁶
  

The application of these principles from Goldman Sachs was fatal to Plaintiffs’ motion for class certification.

In addition to Defendants’ ability to rebut the presumption of reliance by showing a lack of price impact, the Court found, among other things, that Plaintiffs failed to invoke the Affiliated Ute presumption of reliance²⁷  and failed to provide a feasible subclass definition.²⁸  Finally, the Court found that both plaintiffs presented adequacy problems, either having credibility entirely undermined or lacking a fundamental understanding of the case.

While Plaintiff SoCal has since filed a motion for leave to file a renewed motion for class certification, which has been fully briefed, this case provides a recent example of how the principles set forth in Goldman Sachs may be utilized in addressing class certification.

Delaware Court of Chancery Reaffirms that Oversight Liability Requires Bad Faith

Pursuant to the seminal decision of In re Caremark Int’l Inc. Deriv. Litig.²⁹ and its progeny, a claim for breach of the duty of oversight has been predicated on a bad faith failure by fiduciaries to oversee a company’s legal compliance.

Since the Delaware Supreme Court’s 2019 ruling in Marchand v. Barnhill finding a duty of oversight claim survived a motion to dismiss, plaintiffs have aggressively pursued such claims in the Delaware courts. One avenue plaintiffs have pursued is to attempt to expand the types of “mission critical” legal risks that corporate fiduciaries must oversee.³⁰  However, as we have previously discussed, recent decisions from the Court of Chancery reaffirm that oversight liability is predicated on legal risks, not business risks, and remain “possibly the most difficult theory in corporation law upon which a plaintiff might hope to win a judgment.”³¹

In addition to emphasizing that oversight claims are limited to legal risks, the Court of Chancery has recently reiterated that a Caremark claim can only succeed when a plaintiff sets forth factual allegations of bad faith. On this, the burden for a plaintiff pleading bad faith is high, requiring factual allegations that:

• “the directors utterly failed to implement any reporting or information system or controls”;
• “having implemented such a system or controls, [the directors] consciously failed to monitor or oversee its operations”; or
• “directors and officers purposely caused the corporation to break the law in pursuit of greater profits.”³²

Fixing a Caremark claim within the definition of bad faith serves a valuable purpose: “a demanding test of liability in the oversight context is probably beneficial to corporate shareholders as a class, since it makes board service by qualified persons more likely, while continuing to act as a stimulus to good faith performance of duty by such directors.”³³

A recent case emphasizing the difficulty of pleading bad faith is Bricklayers Pension Fund of W. Pennsylvania v. Brinkley (“Bricklayers”).³⁴  In Bricklayers, Vice Chancellor Zurn of the Court of Chancery dismissed claims premised on the alleged failure of oversight by the board of Centene Corporation related to an alleged scheme by four officers to increase their compensation by inaccurately reporting costs associated with a business contract (the “Cost Reporting Scheme”).³⁵ Even though the inaccurate reports, when revealed, required the company to defend and settle regulatory enforcement actions arising from those reports, the Court held that the plaintiff failed to plead particularized facts excusing a demand under both Caremark theories. First, the Board failed to implement and monitor compliance policies and systems to ensure compliance with applicable law, and second, the Board ignored red flags indicating that Centene was not complying with applicable law.³⁶

In particular, the Court held that the allegations in the complaint, and the formal board materials incorporated by reference into it, demonstrated that the Board oversaw and sought improvements to the company’s compliance and board-level reporting systems “for as long as it was aware of the deficiencies at the heart of Plaintiff’s claim.”³⁷ Such active engagement by the Board fell “far short of a showing [of bad faith] that the directors ‘turned a blind eye’ to problems with its reporting systems or knew that Centene effectively had no controls in place.”³⁸ In reaching its decision, the Court distinguished the earlier cases of Marchand³⁹ and Boeing,⁴⁰ each of which painted an “extreme” picture of board-level disregard based on allegations of the failure to implement board-level reporting systems and the failure to respond to red flags that supported a finding of bad faith.⁴¹  The Centene board, in contrast, had compliance and reporting systems in place, including at the board level, and Plaintiff failed to plead any red flags regarding potential illegality from the Cost Reporting Scheme that were both reported to the Centene Board and disregarded.⁴²  Rather, the Court held the Board had reasonably relied on the decisions and recommendations of management regarding compliance issues and regulatory risks.⁴³ Thus, while the company had faced a bad outcome, the Court noted the longstanding principle that “a bad outcome, without more, does not equate to bad faith.”⁴⁴

This case joins a growing list of recent decisions in which the Court of Chancery has reaffirmed the high bar that plaintiffs must still clear to plead a bad faith failure to oversee the company’s legal compliance efforts.