Korean game developers and lawyers are recasting imitation battles around a new problem: artificial intelligence is making it easier to clone not just game ideas, but code, assets and even whole client builds.
Korean game law faces AI cloning pressure
That shift matters because the legal tools that have governed game-copy disputes were built for an era when rivals could borrow a genre, a rule set or a look-and-feel, but not reproduce a working product overnight. As AI lowers the cost of duplication, the economic damage lands faster on smaller studios, while the legal system still moves on a multi-year timetable.
The debate in Seoul is anchored in the evolution of Korean copyright law. For years, courts treated game rules, genres and broad structures as ideas, leaving only concrete expression protected. That began to change with the Supreme Court’s 2019 ruling in the match-3 title “Forest Mania,” which held that even individually unprotectable elements can receive protection if their selection, arrangement and combination reflect creative individuality.
That decision opened the door to a new wave of litigation, including disputes involving NCSoft and Webzen’s “Lineage M” and “R2M,” Nexon and Ironmace’s “Dark and Darker,” and 111 Percent. But the plaintiffs have still struggled to win on copyright alone. Lawyers speaking at the forum said Korean courts remain cautious and typically demand a clear degree of originality that is difficult to show once a genre matures or the disputed mechanics look familiar to the market.
That is why many plaintiffs now lean on Korea’s unfair competition law, specifically the anti-appropriation clause that bars the unauthorized use of another party’s results obtained through substantial investment and effort. In practice, the clause has become the fallback theory in most game lawsuits. A lower court in the 111 Percent dispute accepted that argument even after rejecting copyright infringement, finding that the systematic combination of game rules, UI and progression mechanics had economic value as a protected result.
Still, lawyers at the forum stressed that the clause is not meant to be a catch-all remedy. Courts have been careful to limit it to cases of obvious copying and unfair free-riding, because overuse risks turning a narrow supplement into the main route for protection. The result is a legal framework that protects against blatant imitation, but not quickly or consistently enough to satisfy many studios.
That gap is getting wider as AI enters development. Anchornode chief executive Won Jae-ho said developers are now asking how to survive when a game can be decompiled, copied and uploaded to a Chinese platform in a day. He argued that AI can make it easy to reproduce successful titles nearly verbatim, not just to build “similar” games based on common mechanics. For smaller developers, the problem is less the theory of copying than the speed of enforcement: he said relief often takes one to two years, and sometimes longer, by which time a start-up may already be broken.
The economic stakes are clear. Game development is a hit-driven business, and the value of a creative success is concentrated in the first months after launch. If cloning becomes faster than legal response, the payoff to original development shrinks, deterrence weakens and the industry’s incentives tilt toward larger firms with the capital to absorb litigation costs and document every stage of development.
That concern is not limited to copyright. Gachon University professor Jeong Seong-min said generative AI is already generating a flood of patent applications and will likely push companies toward heavier documentation of design notes and development records as proof against allegations of copying. That may help courts distinguish legitimate inspiration from theft, but it also raises compliance costs that can hit smaller studios hardest and make entry more difficult.
Lawyers also warned that AI creates the opposite problem: developers can accidentally produce games that resemble existing titles without intending to copy them. That raises the need for better similarity testing, clearer records of design intent and perhaps new administrative support to help small firms gather evidence before they go to court. Several speakers argued that administrative relief, investigative support and more accessible IP procedures would matter as much as courtroom doctrine in an AI era.
| Entity | Gains | Losses |
|---|---|---|
| Large game publishers | ▲Stronger legal leverage | ▼Higher compliance costs |
| Indie studios | ▲Faster evidence support | ▼Greater cloning risk |
| AI-assisted developers | ▲Quicker prototyping | ▼Higher infringement exposure |
| Clone operators | ▲Lower copying costs | ▼More legal scrutiny |



