Over two-thirds of Fortune 500 firms and 70% of U.S. initial public offerings are incorporated in Delaware given the historic predictability and consistency of its specialized court system; but sadly times are changing. A controversial recent court decision in a widely watched case involving Cloudflare has resparked fresh fears that the unfortunate corporate exodus out of Delaware will only accelerate from here.

A disclosure before anything else

Cloudflare’s special committee retained us, based on our four decades of recognized governance expertise, to present our widely published research on when companies with dual-class shares outperform and when companies with dual-class shares underperform as the committee deliberated its dual-class share structure. We shared with the committee what we have published for years in academic outlets and in Fortune, as well as elsewhere. This finding is that well-governed, founder-led companies with dual-class shares have, as a group, outperformed other firms in our studies of the entire universe of major firms with such capital structures. At the same time, we found that dual-class structures should expire in instances where they no longer make sense or are abused. We were not asked to opine on the terms of Cloudflare’s own specific proposed recapitalization design, and we did not.

We take no explicit position on Cloudflare’s proposed reclassification. We have no relationship with Vice Chancellor J. Travis Laster, who wrote the opinion, and had no direct dealings with him now or previously. Our glancing role in this matter was just based on a single hour-long presentation before the board. Thus, this piece is not about us. It is about what every Delaware board is now supposed to make of the vague if not vacuous words “good faith.”

How the Cloudflare opinion has resparked fears that the corporate exodus from Delaware will only accelerate

With corporate anchoring fees rivaling personal income taxes as Delaware’s primary source of revenue, Delaware’s wise leaders, including its governor and its legislature, saw this exodus coming, and tried to stem its root cause. In March 2025, Delaware passed Senate Bill 21 (SB 21), the most significant rewrite of its corporate code in a generation, to stop the flow of companies out of the state. SB 21 was generally recognized to have been a direct response to a run of activist rulings from the Court of Chancery, which were widely seen as blatant judicial overreach.

In February 2024, the Court of Chancery voided most of a stockholder agreement at Moelis & Co. as an unlawful constraint on the board, calling into question standard governance arrangements across corporate America. Within four months, the Delaware legislature enacted a new provision, Section 122(18), to authorize exactly those agreements; and the Delaware Supreme Court reversed the ruling and vacated the $6 million fee awarded to the plaintiff’s lawyers.

That same month in 2024, the court ruled that TripAdvisor’s decision to move to Nevada would be judged under “entire fairness,” the most hostile standard in corporate law, because leaving Delaware reduced its directors’ exposure to lawsuits. The Supreme Court reversed that too, holding that ordinary business judgment applies and that the choice between Delaware and Nevada belongs to “legislatures, market participants, and directors,” not to the Court of Chancery. Similarly, the Supreme Court reversed a nearly $200 million original judgment against TC Energy in the Columbia Pipeline case.

No less a luminary than revered former Delaware Chancellor William Chandler, who led the Delaware Court of Chancery for 14 years, testified in support of the passage of SB 21 that “Right now, the corporate market is not feeling good about Delaware.” SB 21 followed, drafted with the help of Chancellor Chandler and former Chief Justice Leo Strine, both widely admired objective legal titans.

Now, Laster’s opinion in the new widely watched Cloudflare case represented the first opportunity to apply the new law, SB 21. The case featured a strong-performing company with dual-class shares, which sought a recapitalization and reclassification to retain the company’s founders and maintain voting control, consistent with a superior track record of sustained outperformance with the stock up ~1800% since going public in 2019.

The spirit of SB 21 is simple: if a board routes a possibly ‘conflicted’ deal through a committee of independent directors with real power to reject it, the deal is considered “kosher.” This means no injunctions, no damages, no trial.

Indeed, Laster’s 142-page opinion concedes the legitimacy of SB 21 and its applicability to this case. He held the directors were independent. He held they were not negligent. He wrote that their process showed “all the outward signs of diligent and careful effort.” By his own account, virtually every requirement the legislature wrote was met by the Cloudflare board.

He enjoined the transaction anyway, on a novel basis. SB 21 does require that the committee act “in good faith,” and Laster’s route was to re-define what good faith entails.

The term “good faith” was not a third level of diligence beyond the duty of care and duty of loyalty, but merely presumed to be the spirit of board actions. Before this, good faith had been previously generally assumed, in deference to independent boards, with nary any heightened review threshold, much less any invocation to enjoin a transaction.

Startlingly, Laster largely set aside, in practice, the existing Delaware definition of bad faith, which is grounded in scienter, or the requirement of showing that individuals acted with knowledge and intent of wrongdoing, as well as the requirement that each board member be assessed individually to determine whether a majority intentionally acted in bad faith.

Instead, he pointed in part to the fact that the board understandably used Signal to communicate about sensitive strategic information they presumably did not want competitors obtaining, irrespective of litigation, and to a CEO who was known to have a hot temper, even though the temperament of the entrepreneur is irrelevant to good faith and typical of creative businesses in technology, media, fashion and the like. None of that amounts to anything even remotely resembling scienter, which is knowing or willful misconduct.

The opinion’s strongest argument should be stated plainly, because boards will have to reckon with it. The court did not rest on any one fact, instead pointing to a fact pattern. It pointed to a junior member of the financial advisers team’s contemporaneous notes about building a “narrative” and securing “gets” for the founders despite those notes being discredited subsequently; to one committee member’s unauthorized conversation telling a founder the committee would give him what he wanted; to two non-committee directors inserting themselves into the negotiation after the founder’s threats; to terms that landed where the founder had predicted; and to messages set to auto-delete after litigation had been threatened, from which the court drew inferences.

From those disconnected, misconstrued scraps, the court inferred that the committee was serving the founders rather than bargaining against them. Reasonable readers can weigh that account differently. Our point is narrower: every item in it is evidence about how the committee negotiated, not about what any individual director intended – and the statute’s word is “good faith,” which Delaware law has always defined by intent.

Laster acknowledged that without what he called the defendants’ “credibility problems,” he “could have concluded” the safe harbor applied. In other words, the statute’s protection now turns on whether a trial judge likes the witnesses. And in this case, the trial judge already judged that the defendants have ‘credibility problems’ without hearing live witnesses testify at trial or sizing them up in person, based only on the paper record of a preliminary injunction hearing.

In the appellate process, usually, trial judges are given deference for factual determinations, based upon their direct personal interactions, but how can any trial judge make any credible factual determinations about the witnesses whom they have never met, did not evaluate in person, and based on only a paper record and remote deposition transcripts?

What “good faith” now means for boards

If the Laster opinion is allowed to stand as precedent, the practical meaning of “good faith” for a Delaware board is this: there is no checklist, no process, no roster of advisers and no set of minutes that can secure it in advance. A committee that meets every structural requirement the legislature wrote can still be sent to trial because a single judge, reading advisers’ notes and text messages months after the fact, concludes that its heart was not in it. Boards will have to put up with that, deal by deal and judge by judge, until the Supreme Court or the legislature says otherwise.

Consider one concrete example of what boards are now being asked to anticipate. The opinion faults Cloudflare’s committee for not obtaining a fairness opinion and for not quantifying the value of the benefits the founders would receive — while conceding, in its own footnote, that “Delaware courts have stated that there is no duty to obtain a fairness opinion,” and that Delaware decisions “have shown greater deference to a decision not to secure a fairness opinion, even for purposes of a controller recapitalization.” That concession is not a technicality.

Fairness opinions are strictly instruments of mergers and divestitures, where there is a price to be fair about. They are never obtained for recapitalizations, and no investment bank can credibly opine on the “fairness” of extending founder control against an imaginary alternative that was never on the table — which is why Delaware’s own precedents, from the NRG Yield reclassification onward, have rejected the argument that one is required. A board that reads this opinion learns that it can be faulted for omitting a step the law does not require and which nobody in the market can reasonably perform.

And that is why the Cloudflare decision matters far beyond one company’s share structure, and even beyond dual class shares. Every board in America that is still incorporated in Delaware, especially companies with strong founders and controllers, now has the same question in front of it: if this fair, independent process was not enough, what would be enough to satisfy Delaware judges on what ‘good faith’ entails?

A board that does not know what “good faith” means in practice has no safe harbor at all; it has a lottery ticket with the odds stacked against it. Delaware’s entire value proposition, for a century, was that it was the one jurisdiction where boards could know the rules before they acted, in a predictable and consistent way. That predictability and consistency have now gone up in thin air.

Boards are not sentimental about their legal domicile. The forty-four reincorporation proposals this year, up from seventeen two years ago, represent a natural response to this heightened Delaware risk. Delaware’s legislature understood that in March of 2025, when it passed SB 21 to put a floor under that calculation. Vice Chancellor Laster has now pulled the floor out again for all substantive purposes.

None of this is an indictment of the Delaware Court of Chancery as an institution. In fact, the court’s other distinguished judges have, if anything, been saying the quiet part aloud. For example, highly respected Vice Chancellor Lori Will — who certified the SB 21 constitutional questions to the Supreme Court within weeks in 2025 so that “courts, corporations, litigants, and transaction planners” would have clarity and predictability — published a paper last fall titled Back to Basics: Delaware’s Genius of Simplicity, at the UPenn Law School and forthcoming in the Journal of Corporation Law. Its thesis is that Delaware’s strength lies in “simple, enduring principles,” that “modern pressures invite complexity and the expansion of fiduciary theories [with] clarity put at risk,” and that “preserving stability and judicial restraint is essential to sustaining the predictability that has long defined our law.”

To be fair, this latest ruling is interim, and the Delaware Supreme Court will almost certainly get its turn. Yet this will involve more rounds of litigation. Practically, the unfortunate reality for Delaware is that the damage of each reversal is cumulative, not corrected.

Every company watching the Cloudflare saga play out in Delaware, especially companies with strong founders and controllers, has to wonder whether it is time to reincorporate in Texas or Nevada. The nation’s companies in competitive technology sectors cannot help but note that even if the decision is reversed, such a correction would still mean that Cloudflare will have lost well over a year – in the best case – in implementing its capital structure.

Hundreds of remaining Delaware-based companies are not shrugging this off as “water over the dam” safe to ignore. As historian David McCullough wrote of the tragic inaction ahead of the catastrophic 1889 breach of the Johnstown, Pennsylvania dam, people ignored a growing stream over the structure’s top. The growing over-the-top Chancery Court rulings may well lead to a stream of corporate exits.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.