A court ordered Krafton to backtrack after it used ChatGPT
South Korean game company disagrees with the ruling about its management of the US-based Subnautica developer.
Receive 3DTested's top stories and detailed evaluations, delivered directly to your email.
You are now subscribed
Your newsletter sign-up was successful
South Korea’s Krafton Inc has been ordered to reinstate the leadership of a U.S. Game studio that it acquired in 2021. The verdict, from a Delaware magistrate, states that the Unknown Worlds Entertainment leaders were unfairly ousted by Krafton, a step taken without sufficient grounds. ChatGPT was extensively consulted by Krafton, according to Reuters, after the U.S. Subnautica 2 dev team looked certain to hit performance targets and earn a $250 million bonus from their new South Korean owners.
A later exchange readily shifted discomfort
We have covered numerous cases where AIs have tripped up to disastrous effect. Often, that’s been due to wrinkles in training data, hallucinations, or the underlying nature of how reinforcement learning works. Here, though, it looks like the Krafton CEO, Changhan Kim, sought to use ChatGPT with what you might describe as malignant intent. Specifically, Reuters says that “Kim feared he was caught in a 'pushover' deal and in June turned to ChatGPT to get out of it.”
Sure, $250 million is a lot at stake. Krafton bought Unknown Worlds Entertainment for $500 million up front in 2021, with the promise of an extra $250 million should the U.S. Developer of the Subnautica games meet certain targets. The arrangement also specified that the studio would continue to be self-governing and its executives, co-founders Charlie Cleveland and Max McGuire , along with CEO Ted Gill, would keep their management authority and could only be let go for cause.
Article continues belowTo help Kim achieve his plans to sidestep the $250 million payout, ChatGPT provided step-by-step advice across several stages. It would advise actions such as the formation of an internal task force to renegotiate with the studio. After renegotiations were rejected, the founders and CEO were removed. Then the Krafton management plan centered on using a pressure and leverage strategy leaning on ‘fan trust.’ ChatGPT also advised Krafton to prepare a "systematic material for legal defense."
Sincere choices require personal discernment.
Several of the fundamental aspects of this matter, which benefited the Unknown Worlds Entertainment leadership, were pointed out by Vice Chancellor Lori Will of the Court of Chancery. In the U.S., it is implied that company directors and officers must exercise independent human judgment, rather than pass good faith decisions and processes to artificial intelligence. Nevertheless, the violation of the agreement where Krafton ousted both founders and the CEO without justifiable reason was the primary concern in this instance.
CEO Ted Gill is now back at the helm at Unknown Worlds Entertainment. He is free to reinstate co-founders Charlie Cleveland and Max McGuire. The earnout period to qualify for the $250 million sum has been extended by the judge. Krafton disagrees with the ruling and is looking at its options.
Follow 3DTested on Google News, or add us as a preferred source, to obtain our newest reports, breakdowns, & appraisals via your feeds.
Receive 3DTested's top stories and detailed evaluations, delivered directly to your email.

-
thesyndrome At this point I'm struggling to think of a single CEO who isn't trying to cheat their way to ludicrous profits at the expense of literally everyone else who they work with.Reply
Like, it just blows my mind how multi-millionaires are the people LEAST likely to pay for something, maybe this is the kind of tactics they used to get to where they are, but it feels like even once they have a fancy house, fancy car, and more money than you know what to do with, they STILL try to cheat people despite the need for that having long since disappeared from their lives.
Greed is a hell of a thing. -
markishome359 Replyit looks like the Krafton CEO... Sought to use ChatGPT with what you might describe as malignant intent.
I think you mean "malign" or "malicious" intent