Skip to content
Submit incident
AII-1002-59563
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

Krafton CEO used ChatGPT to plan removal of studio founders to avoid payout

Krafton's CEO used ChatGPT to plan the removal of Unknown Worlds Entertainment's founders to avoid paying a $250 million earnout bonus owed under their acquisition agreement. A Delaware court found the scheme violated the acquisition contract and ordered Krafton to reverse the removal and reinstate the studio's CEO.

2025 Β· Krafton Β· Governance, accountability and labour
2025
AII-1002-59563
Quick facts
Category
Governance, accountability and labour
Date
2025
Drafted by
Raidu Incident Detection

What happened

Krafton CEO Changhan Kim consulted ChatGPT for advice on how to remove the founders of Unknown Worlds Entertainment and avoid paying a $250 million earnout bonus owed under the 2021 acquisition agreement.

Based on ChatGPT's guidance, Krafton executives removed the studio's CEO and cofounders from control of the company. The founders were fired and Krafton took control of publishing rights for Subnautica 2, delaying the game's release.

A Delaware Court of Chancery judge found that Kim had used ChatGPT to engineer a corporate 'takeover' strategy and ordered Krafton to reverse its actions and reinstate the CEO of Unknown Worlds Entertainment.

Reported impact

The founders lost operational control of the studio and faced reputational harm. Unknown Worlds staff and players were affected by the delay in Subnautica 2's release.

The case raised broader concerns about executive accountability and how generative AI can be used to optimise tactics that may undermine contracts, workers, and investors.

Index analysis

The incident highlights a transparency problem: Krafton allegedly relied on internal AI-assisted planning for corporate strategy and refused to produce the ChatGPT conversations when pressed, making it difficult for courts and affected parties to understand decision-making.

For policymakers, the case demonstrates the need for clearer record-keeping, disclosure, and governance rules when companies use AI in decisions with legal and financial consequences.

Governance controls

Krafton's internal controls failed to prevent the CEO from using AI to plan actions that breached contractual obligations to acquired company founders.

Classification

Category
Governance, accountability and labour
Industry
Arts, entertainment and recreation
Harm status
Actual harm
Intent
Intentional
Timing
Unknown
Harm types
Economic or property, Reputational
Harmed
Business, Workers
Failure modes
Agent misaction
Consequences
Litigation
AI systems
ChatGPT

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident