Submit incident
Documented

ChatGPT Drafted the Strategy That Led to Three Wrongful Terminations, and a Delaware Court Tied Them Together

July 1, 2025
Curated by Team Raidu · Reviewed by Shiva Ganesh
aiid:1449View source ↗
LinkedInX

What happened

Krafton, the South Korean publisher behind the PUBG franchise, acquired Unknown Worlds Entertainment, the studio that made Subnautica. Like most acquisitions of privately held game studios, the deal carried earnout provisions: performance-based payments owed to the studio's leadership team contingent on hitting targets after the deal closed. When questions arose about whether those obligations could be avoided by removing the executives they were tied to, Krafton's CEO received advice making clear they could not. Firing the Unknown Worlds leaders would not make the earnout go away.

The CEO turned to ChatGPT anyway. According to a Delaware Chancery Court opinion issued in July 2025, Krafton sought from the AI a "no-deal" strategy, a set of steps for escaping the obligations the deal had created. The court found that Krafton then followed most of what ChatGPT recommended. What had started as an internal business problem became a documented sequence: a query to a commercial AI tool, a list of recommended actions, and a company that acted on them.

The actions the court tied to that strategy were not minor course corrections. Krafton locked Unknown Worlds out of control over its own publishing decisions. The company issued public communications about the studio unilaterally, without the involvement of studio leadership. And Krafton terminated three executives who had run the studio: Ted Gill, Charlie Cleveland, and Max McGuire. All three had been central to Unknown Worlds' operations and to the earnout structure the acquisition created.

The Delaware Chancery opinion treated the ChatGPT consultation as evidence of intent. The court's analysis connected the AI-generated recommendations directly to the conduct that followed, characterizing the terminations and the seizure of studio control as parts of a coordinated campaign rather than independent management decisions. That framing placed the entire sequence inside a single legal narrative, and the ChatGPT exchange sat near its start. Courts evaluating intent rarely get to examine this kind of artifact, a timestamped record of a strategy being assembled in real time.

The case surfaces something that goes beyond one acquisition dispute. Commercial AI chat tools leave records. A consultation that feels like informal brainstorming produces a timestamped log, one that can be retrieved, subpoenaed, and placed in a court filing. Nothing in standard corporate governance practice currently accounts for how that record differs from a whiteboard session or a verbal briefing. A provable record of what a system produced, and what organizational actions followed from it, can shift a legal narrative in ways neither party anticipated when the chat window opened.

Reported impact

Affected parties
Not publicly disclosed
Harm type
Not publicly disclosed
Scale
Not publicly disclosed
Financial impact
Not publicly disclosed
Regulatory action
Not publicly disclosed

Classification

Organization
Not publicly disclosed
AI system
Not publicly disclosed
Industry
Not publicly disclosed
Country
Not publicly disclosed
Provider
Not publicly disclosed
Incident type
Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

  • No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.

AI Incident Database
Also catalogued in
ChatGPT Drafted the Strategy That Led to Three Wrongful Terminations, and a Delaware Court Tied Them Together
2025-07-01