On July 16, 2025, for perhaps the first time in the history of Delaware corporate law, a Caremark claim involving a public company director went to trial. Meta stockholders brought an $8 billion claim against the company’s directors and officers, arguing that they failed to oversee Meta’s compliance with its privacy obligations. During the proceedings, Vice Chancellor Laster described the alleged conduct as occurring on a “truly colossal scale.” By the second day of trial, the parties announced a confidential settlement, ending the proceedings before key executives could take the stand.
The eleventh-hour Meta settlement captures a paradox in modern oversight litigation: Caremark claims have become increasingly common and generate enormous boardroom anxiety, yet they almost never reach trial, let alone result in personal liability. In the three decades since Chancellor Allen characterized oversight liability as “possibly the most difficult theory in corporation law upon which a plaintiff might hope to win a judgment,” Caremark has evolved from a narrow doctrinal anomaly into the foundation of a compliance apparatus whose annual cost to companies is now hundreds of billions of dollars. Law firm alerts herald each new decision as evidence of a “new Caremark era” while directors are warned of unprecedented personal exposure for failing to implement board-level monitoring systems.
But what if this narrative is wrong?
In a new article, I present acomprehensive empirical account of Caremark claims. The study employs a hand collected dataset of Delaware oversight decisions from the doctrine’s inception in 1996 through 2024. It draws on judicial opinions and, where available, the underlying pleadings and hearing transcripts. To identify what kinds of claims the courts confront and how they respond, I systematically coded each case for the theory of liability pled, the event that triggered litigation, procedural posture, outcome, and judicial rhetoric. The results reveal a striking gap between Caremark’s symbolic prominence and its doctrinal modesty, a phenomenon the article calls the “oversizing” of Caremark.
Liability Rarer Than the Narrative Suggests
The empirical picture is considerably more restrained than the prevailing account of Caremark suggests. Nearly every decision in the dataset (96.3%) was rendered on a motion to dismiss, and only one oversight case in the doctrine’s history has produced a post-trial opinion. Because so few cases proceed beyond the pleading stage, the question is whether plaintiffs can survive a motion to dismiss. On that measure, Delaware courts found allegations sufficient in 25.6% of cases, a rate that is modest, though notably higher than the near-zero probability often assumed.
While the first finding concerns the level of success, the study’s second finding concerns the disconnect between litigation volume and outcomes. Although the annual number of Caremark decisions nearly doubled after Marchand, plaintiffs’ pleading-stage success rate remained essentially flat (25.0% after Marchand compared with 26.1% before). Marchand thus appears to have encouraged more filings without fundamentally altering plaintiffs’ odds of success. The perceived heightening of oversight standards, in other words, shows up in the volume of litigation and in the surrounding rhetoric, but not in case outcomes.
The Focus on Bad Actors
The aggregate success rate, however, tells only part of the story. To understand what Caremark actually targets, it is necessary to look more closely at the theories of liability that plaintiffs plead and the claims that courts allow to proceed. The conventional account credits Caremark with creating a duty to implement information and reporting systems (Information Systems claims). Yet these claims appear in fewer than half of all cases and rarely succeed. Even when plaintiffs prevail on the Information Systems claim, the court frequently resolves the motion on alternative grounds, thus acknowledging the Information-Systems theory without meaningfully applying it.
The data reveal that the docket is instead dominated by Red Flag claims—allegations that directors ignored clear warning signs of misconduct. These claims predate Caremark and trace their lineage to Graham v. Allis-Chalmers. Although Red Flag claims appear in 89% of cases, their prevalence does not translate into success. When pled alone, without any accompanying theory, they failed in all 28 cases. This result suggests that Delaware courts remain deeply skeptical of allegations that a board should have reacted differently to ambiguous warnings.
The winner is a theory that has received far less attention: Massey-style claims, which allege that directors knowingly caused the corporation to pursue an illegal business strategy. Although these claims appear in barely a quarter of cases, they succeed at a 45.5% rate, which is the highest of any oversight theory, and roughly twice that of the dataset’s baseline. This contrast suggests that Delaware courts are more receptive to claims targeting bad actors than to claims challenging bad systems.
Indeed, across nearly three decades of litigation, successful oversight claims cluster around three narrow scenarios: (1) monitoring systems that were deliberately designed or knowingly maintained to be ineffective; (2) directors’ failure to oversee misconduct by the corporation’s own fiduciaries; and (3) the deliberate adoption of illegal business practices as a matter of corporate strategy. None of these successful theories reflects a broad judicial endorsement of liability for inadequate proactive monitoring. Instead, each involves a conscious disregard of legal boundaries rather than an imperfect compliance architecture.
When Narrative Outruns Doctrine
If Caremark liability is so constrained, then why all the boardroom anxiety?
The answer lies in the environment surrounding it. Law firm client alerts routinely translate each new oversight decision into urgent calls for governance reform. Judicial opinions have become increasingly critical in tone even as dismissal rates remain largely unchanged. Likewise, a mature and lucrative compliance industry reinforces these messages. Taken together, this environment creates a self-reinforcing feedback loop of expansive proactive duties that extends well beyond the doctrine’s modest origins and application.
That does not make the narrative inconsequential. Quite the opposite. Caremark’s greatest influence lies in shaping governance norms rather than expanding liability. Delaware courts continue to impose demanding thresholds for liability while simultaneously producing a stream of judicial commentary that, amplified by law firms and compliance professionals, reshapes boardroom expectations. Caremark therefore operates less as a liability doctrine and more as a source of governance guidance that steers boardroom behavior despite its limited traction in the courtroom.
Lessons for Boards, Courts, and Compliance
For directors, the study offers both reassurance and a warning. The low incidence of Caremark liability should temper fears of personal exposure for good-faith oversight efforts. At the same time, the concentration of successful claims around deliberate misconduct demands heightened vigilance when commercial pressures encourage unlawful conduct. Delaware courts tolerate imperfect monitoring, but they are far less forgiving of conscious lawbreaking dressed up as business strategy.
The implications extend beyond directors. For the compliance industry, the data suggest a misallocation of resources. Billions of dollars are spent annually on elaborate monitoring architecture, which may satisfy the prevailing Caremarknarrative while doing little to address the failures that actually generate liability. For the judiciary, the findings identify a choice—either to develop Information Systems theory into a meaningful doctrinal framework or to acknowledge that Delaware corporate law continues to rely primarily on reactive oversight.
Yehonatan Shiman is an assistant professor at Ono Academic College, Faculty of Law, and the director of the law program at Ono International School. This post is based on his recent article, “Oversizing Caremark: An Empirical Analysis,” available here.