Let me be direct with you. Your AI conversations are not private. They are not protected. And they can be used against your company in court.

On March 16, 2026, the Delaware Court of Chancery made that crystal clear in Fortis Advisors v. Krafton. Every board director needs to understand what happened in this case and what it means for how your organization governs AI use right now.

What the Court Found

Krafton, a South Korean gaming conglomerate, acquired Unknown Worlds Entertainment, the studio behind the popular game Subnautica, in 2021 for $500 million upfront. The deal included up to $250 million in contingent earnout payments. The founders and CEO Ted Gill received contractual guarantees of operational control and protection from termination without cause.

When internal projections showed the upcoming release of Subnautica 2 would trigger a payment approaching $250 million, Krafton’s CEO CH Kim wanted out. His own legal team told him a for-cause termination would not eliminate the obligation and would expose Krafton to litigation. He ignored that advice and turned to ChatGPT instead.

The AI chatbot gave him a detailed corporate takeover strategy, including a pressure-and-leverage package, specific control points to seize, and a two-handed approach combining legal tactics with softer incentives. Kim followed the playbook. Krafton locked Unknown Worlds out of its Steam publishing platform, posted unauthorized messages on the studio’s website, and terminated all three executives.

Then the litigation began. The court quoted those ChatGPT conversations at length. It found every justification Krafton offered for the terminations was pretextual. CEO Ted Gill was reinstated with full operational authority. The earnout period was extended by 258 days to replace the time Krafton’s breach had consumed.

Kim also deleted some of those AI logs after litigation began. Don’t get fooled into thinking deletion solves the problem. The court noted the spoliation, and it will cost Krafton more in the damages phase ahead. Much more.

What Your Board Needs to Understand About AI Governance

Here is the governance lesson every board needs to take from this case.

AI conversations are discoverable. When an executive types a strategy into ChatGPT and receives a detailed tactical plan in response, that exchange can be subpoenaed, entered into evidence, and quoted by a judge in a published ruling. Attorney-client privilege does not extend to a publicly available AI tool. That is not a future concern. That is what happened in a Delaware courtroom in 2026.

Ask yourself this: Does your executive team understand that? If the answer is anything other than an unqualified yes, you have a governance gap that needs to be closed today.

The Krafton case is the first major ruling in which AI-generated strategy conversations were treated as direct evidence of bad faith and pretext at trial. It will not be the last. Courts are not developing new frameworks for AI. They are applying existing legal frameworks to new tools, and the results look exactly like what happened to Krafton.

This is precisely the kind of reputational and legal exposure that proactive crisis management is designed to prevent. A board that governs well does not react to these situations. It builds the systems that make them unlikely.

The Three Governance Questions Your Board Must Answer

Does your organization have an AI governance policy that specifically addresses how executives document, retain, and handle AI-generated strategy materials? This is not a cybersecurity question alone. It is an evidentiary risk question, and most companies have not addressed it.

Are your executives treating AI conversations with the same discipline they apply to legal memos, board communications, and formal correspondence? If the answer is no, both the behavior and the written policy need to change now.

Is your board reviewing AI use as part of its risk oversight function? The Harvard Law School Forum on Corporate Governance [OUTBOUND LINK] identified this case as a direct warning for boards: AI tool use by executives sits squarely within the risk oversight mandate. If it is not on your risk committee’s agenda, put it there.

Not sure where your organization actually stands? Take our Crisis Readiness Scorecard to identify where your vulnerabilities are before they cost you credibility or legal exposure.

What Good AI Governance Looks Like in Practice

The goal is not to prohibit executives from using AI tools. These tools create real value, and your team should use them. The goal is to govern them with the same seriousness you apply to any written record that could surface in litigation.

That means a written policy covering three things. First, AI tools may be used for strategic planning around sensitive matters. Second, how AI-generated materials should be retained or disposed of, consistent with your document retention policy. Third, legal counsel needs to be involved before AI tools are used to develop a strategy around contractual or legal questions.

It also means training. Executives who would not type a litigation strategy into an email without first calling their attorney need to understand that typing it into ChatGPT carries the same risk. In some ways, the exposure is greater because the arguments for privilege are weaker.

Think of it this way: if you would not want a judge reading it in a published opinion, do not type it into an AI chatbot without governance in place.

Boards that do this work proactively avoid situations like Krafton’s. Boards that do not manage the fallout. Narrative control and reputation protection become vastly harder once the AI logs are already in evidence.

The Reputational Damage Lasts Beyond the Courtroom

Here is the part most boards miss entirely. The Krafton ruling is a published opinion. It is publicly searchable. It names the CEO, quotes his AI conversations, and describes the strategy he designed to avoid a contractual obligation.

That record does not disappear when the litigation ends. It lives in legal databases, news archives, and search results. Every future partner, investor, or acquirer who searches Krafton’s name finds this case.

Your board’s job is not just to avoid losing in court. It is to protect your organization’s reputation before a ruling like this becomes the defining public record of how your leadership makes decisions.

The Krafton ruling does not change the law. It applies existing law to a new tool, and the results are exactly what you would expect. Your board’s job is to make sure your executives understand that before the next lawsuit, not after.

If your board is not yet talking about AI governance, this is where that conversation starts. Contact Ethia Strategies to start building the governance and communication framework your organization needs now.