Delaware Court Hammers Earnout Evasion
There is a reason that many M&A attorneys call an earnout (a potential post-closing payment tying part of the purchase price to future performance targets) the “litigator’s invitation to the party.” That reason was illustrated in a recent Delaware Court of Chancery decision. Fortis Advisors, LLC v. Krafton, Inc., C.A. No. 2025-0805-LWW (Del. Ch. Mar. 16, 2026), is a lesson in the consequences of trying to circumvent an earnout provision. The case was also another decision showing that artificial intelligence-based research, even about legal strategy, is discoverable.
What Happened
In 2021, South Korean gaming company Krafton acquired Unknown Worlds
Entertainment—the studio behind the hit game Subnautica—for $500 million up front, plus up to $250 million in earnout payments tied to revenue from the sequel, Subnautica 2. The deal guaranteed that the founders would retain operational control of the studio, and could be terminated only “for cause.”
By mid-2025, Krafton’s own projections showed the game was on track to trigger a payout exceeding $190 million. Fearing that this level of payout was more than they had bargained for, Krafton attempted to block the payout and subsequently fired the founders, citing a “lack of game readiness” as the cause, then later claiming that the founders had abandoned their roles and improperly downloaded company data. Krafton’s corporate development team had advised that this course of action had the potential to invite litigation. Krafton executives, however, ignored those warnings and instead decided to consult an AI chatbot. That service generated a multi-step strategy for renegotiating or circumventing the obligation, part of which included firing the founders and delaying the game’s launch.
What the Court Did
The court found that the founders had not engaged in any behavior that would have
satisfied the “for cause” standard which was necessary for their removal and that Krafton therefore breached the acquisition agreement by terminating the key employees without valid cause and seizing operational control of the studio. A court finding that a breach occurred is not news. The court’s remedy, however, is what sets this case apart: the court ordered the CEO reinstated, returned operational control to the CEO, and extended the earnout testing period by 258 days, the exact duration of his wrongful ouster, which pushed the earnout performance deadline to September 15, 2026. Rather than forcing the sellers to prove hypothetical revenues in a damages trial, the judge preserved their opportunity to earn the contingent purchase price on the original terms.
Lessons for Buyers and Sellers
For sellers: Delaware courts are apparently willing to go beyond a cash payment as
compensation for damages, with the justification being that in some cases, such as Fortis, monetary damages cannot provide complete relief because said damages would be based on a hypothetical scenario. If a buyer attempts to block a seller’s earnout through illegitimate means, the court may give the seller back the time and opportunity that was taken from it.
For buyers: Bad faith efforts to dodge a contractually guaranteed earnout carry
extraordinary risk, especially when coupled with the unjustified firing of founding
employees. Fortis also reemphasizes the point that internal communications, strategy memos and, as this case demonstrates, AI chatbot transcripts, can all become evidence of bad faith, which is not necessarily a new revelation, but is important to note just the same.
For both sides: If the earnout depends on key people staying involved, make sure the agreement spells out what happens if they’re removed. If the buyer has discretion over the business, define its limits. Earnout provisions deserve careful attention before the deal closes, not after. And, once again, even though it was the buyers that were on the losing end of this lawsuit due to their reliance on an AI chatbot for legal advice, the lesson that AI is not yet at the point where it is qualified to provide such advice is one that can certainly apply to both sides, and should be noted accordingly.
There is yet another cautionary takeaway from this case: research done with AI is discoverable. The court specifically noted that Krafton’s CEO “consulted an artificial
intelligence chatbot to contrive a corporate ‘takeover’ strategy.” Had he picked up the phone and called a lawyer, that conversation would have been protected by attorney-client privilege and therefore outside the bounds of discovery. Instead, his chatbot history produced a fully discoverable, searchable record of every prompt and every response, laying out the entire bad faith attempt, from start to finish. That chat log became some of the most damning evidence in the case. And the advice itself was bad: the AI generated a step-by-step plan that walked the CEO straight into a breach of contract. A good lawyer would have recommended that he stop. The chatbot told him how to proceed.
Earnouts can be a smart way to bridge a valuation gap. But as this case reminds us, they are also the litigator’s invitation to the party and, in Delaware, the court is more than willing to RSVP.
