On August 13, 2026, the Delaware Court of Chancery dismissed derivative claims against Boeing's current and former directors in In re The Boeing Co. Derivative Litigation (C.A. No. 2024-1210-MTZ). The case arose from the January 2024 midair door plug failure on a Boeing 737-9 MAX, four years after the company rebuilt its safety oversight following the 2018 and 2019 737 MAX crashes and the $237.5 million derivative settlement that followed them. Vice Chancellor Morgan T. Zurn held that plaintiffs failed to plead the bad faith a Caremark claim requires, even though the incident the suit was built on had actually occurred.
Strategic Significance
The ruling matters less for what it says about Boeing than for what it confirms about the standard every board will eventually face on AI oversight. Boeing's directors survived because the record showed a functioning system, not a clean outcome: an aerospace safety committee of independent directors with relevant expertise, a dedicated safety organization reporting directly to that committee and to the chief engineer, and management safety briefings delivered at every board meeting. The Court rejected the plaintiffs' argument that the sheer volume of safety reporting was itself evidence of an ignored red flag, calling that reporting what it was: diligence, not disloyalty. It also required that any red flag be “sufficiently similar” to the incident that followed, not a general risk raised in passing.
The incident still happened. The board still won. The distinction the Court drew was between an architecture that receives, escalates, and responds to risk information, and one that merely exists on paper. That distinction is the entire game for AI oversight right now, and no Delaware court has yet ruled on a Caremark claim built on an AI failure.
FRICTION POINT: Most boards currently treat AI governance as a stated policy or a committee charter amendment, not as the standing, reporting, responsive architecture Boeing's safety committee represents. A policy on file generates no minutes, no escalations, and no management responses to review. Under the standard this ruling reaffirms, a policy is not what would keep a board out of a books-and-records demand, let alone survive a Caremark claim once one is filed.
Governing Implication
This is the Declarative Board Failure Pattern inverted and proven in a courtroom instead of asserted in a boardroom. Boeing's directors did not declare a commitment to safety and point to the 737 MAX crashes as evidence the message had landed. They built a committee with named members, a reporting line from a specific executive function, and a standing agenda item, and they kept building after the January 2024 incident rather than treating the earlier reforms as sufficient. That is what Delaware rewarded: not the absence of failure, but a system that responded to failure in real time and left a record of having done so.
Boards moving AI oversight from a stated intention to a governed function should read this decision as a preview of what that function will need to look like before the first AI-related Caremark claim is filed, not after. The evidentiary bar will not be novel when it arrives. It will be this one, applied to a new risk category.
Action Point
At the next board meeting, ask whether the AI oversight function has the three things Boeing's safety committee had in place at the time of the January 2024 incident: a standing committee with members who can speak to the specific risk, a reporting line from a named executive function directly to that committee, and minutes showing the committee asked follow-up questions the last time it received a report. Any of the three that is missing is the gap to close before the next filing cycle, not the next piece of litigation.