On Thursday, November 5, 2026, the Seoul Central District Court held the eighth hearing in the copyright and unfair-competition suit that KBS, MBC and SBS brought against Naver over alleged use of their news articles to train generative AI, 기자협회보 (the Journalists Association of Korea) and DigitalToday report.
The broadcasters' statements are allegations. Naver's statements are defenses. No court has ruled on the claims. The sources reviewed report no judgment.
What press say the court noted on claimed damages
기자협회보 reports that Civil Division 63 of the Seoul Central District Court reviewed the plaintiffs' briefs and said the broadcasters had each specified alleged damages of at least 41 billion won and at most 47.7 billion won, based on news production costs. The same report says the three started the case seeking 200 million won each (600 million won in total) and that adding those later figures across the three plaintiffs would put the combined claimed minimum at about 120 billion won (1,200억).
Those amounts are the plaintiffs' claimed figures as stated in court, not a finding of liability or of actual loss. 기자협회보 also notes that the early 200 million won figure per plaintiff was set so the case would clear the 500 million won (5억) threshold for a three-judge panel, with the parties planning to raise the claim later. The November 5 briefs are that later specification, according to the same report. Separately, MLex reported on November 5 that a lawyer for the broadcasters said they were seeking a combined 15 billion won from Naver and its affiliate Naver Cloud. That figure is far lower than the per-plaintiff amounts 기자협회보 attributes to the court, and the Korean hearing reports reviewed here name Naver, not Naver Cloud, as the defendant in court. We could not reconcile the two figures from the available sources, so neither is treated as an established amount of loss.
DigitalToday says the three filed in January 2025 seeking damages for copyright infringement and Unfair Competition Prevention Act violations, plus an order barring AI training on their articles. An MLex summary dated January 14, 2025, also describes the suit as filed that month. It says they allege Naver used the articles to train HyperCLOVA and HyperCLOVA X. That use remains an allegation. DigitalToday frames the core disputes as the scope of copyright protection for news articles and whether existing news-content partnership terms already authorized AI-training use.
SBS News names the panel as Civil Division 63, a three-judge panel, under Presiding Judge Lee Gyu-young.
What the parties argued
DigitalToday and iNews24 say the live issues included whether news articles are protected under copyright law, whether using them for AI training is misappropriation of achievements or unauthorized data use under the Unfair Competition Prevention Act, whether that use is fair use, and what amount of damages, if any, would be appropriate.
The broadcasters told the court, as quoted by DigitalToday: "뉴스 기사를 생성형 AI 학습에 이용하는 것은 약관 적용 당시 예측 가능성이 없었고 이에 대한 대가도 지급된 바 없다." SBS reports they also said that reading the existing partnership terms to cover AI training would make a comprehensive license that is excessively unfavorable to them and therefore void.
기자협회보 quotes their counsel on fair use: "생성형 AI 학습이라는 것은 기사 원문을 정답지로 삼아 그 표현을 재현하도록 모델을 반복적으로 조정하는 과정이고 그 결과로 기사의 표현적 내용이 모델에 남게 된다. 그렇기 때문에 학습은 그 자체로 표현적 비용이고 따라서 변형적 이용 주장은 원고들에게 유리하게 판단돼야 한다." Counsel also asked the court to weigh the time and cost of producing each article, and said Supreme Court guidance makes the effect on the license market the most important fair-use factor.
Naver, as quoted by 기자협회보 in court, said: "원고들이 주장하고 있는 기사들이 대부분 시사 보도라는 점은 저희가 소송 초기부터 말씀드린 바 있다. 저작권법상 보호받을 수 없는 콘텐츠." SBS and iNews24 report Naver also said AlphaGo and ChatGPT already existed when the terms were signed, so AI training use was foreseeable. 기자협회보 says Naver pointed to a 2018 special agreement it says was drawn for AI platform services, and argued that open access to the articles means they are not the kind of data that unfair-competition rules protect. The same report quotes Naver saying it had already paid under the contracts and that "지금 저희가 AI 없이 살 수 없는 세상에 살고 있다. 그런데 이런 경쟁을 불가능하게 만드는 청구를 인용한다는 것은 굉장히 부당하다고 생각된다." Those lines are Naver's courtroom argument, not findings.
The Korean Broadcasters Association statement
After the hearing, the Korean Broadcasters Association said, as quoted by iNews24: "국내 최대 AI 기업인 네이버가 뉴스 콘텐츠는 저작권의 보호 대상이 아니라는 상식 수준을 벗어난 주장으로 일관하고 있다." It also said it deeply regrets what it called disparagement of the copyright value of news, and that if Naver's argument were accepted, copyright in all news could be denied, so copyright holders as a whole need a joint response. SBS carries the same statement. That is the association's view, not a court finding.
Next hearing and why it was added
기자협회보 reports the ninth hearing is set for December 1, 2026, at 2:15 p.m., and that it is expected to close arguments. The same report says the eighth hearing was expected to be the close, but the court set one more date after issues over translation of additional evidence and unpaid filing fees. If a judgment date is set at that hearing, 기자협회보 says a first-instance ruling could come in late December 2026 or early 2027. DigitalToday, SBS and iNews24 also say December 1 is expected to end oral argument.
No ruling is reported in the sources reviewed.
Why the hearing matters for AI training disputes
기자협회보 calls the case South Korea's first AI copyright suit between news organizations and an AI company. It is now past the early claim stage and into named damage figures, with a closing hearing already calendared. The fight turns on three points the outlets name: whether broadcast news counts as protectable expression or unprotected current-affairs reporting, whether older news-partnership terms cover generative AI training, and whether training is fair use under Korean law as the parties argue it.
None of those points is settled in the reporting reviewed here. The November 5 hearing put the size of the plaintiffs' claimed loss on the record and left the legal theories for the December 1 close. Readers who follow publisher and AI-lab fights elsewhere can compare process, not outcomes: the Like Company advocate-general opinion on Gemini and press rights, the Concord lyrics claim against Anthropic, and Anthropic's reported Associated Press licensing deal. Those matters use different laws and facts. None of them decides the Seoul case, and none of the Korean outlets cited here treats a foreign ruling as binding on Naver or the broadcasters.
Sources
- 기자협회보: AI 저작권 소송…지상파 3사 "네이버 끼친 손해액 각 400억대" (November 6, 2026)
- DigitalToday: AI 학습에 쓰인 뉴스 저작권 놓고 지상파 3사 vs 네이버 법정서 공방 (November 5, 2026)
- SBS News: 한국방송협회 "국내 최대 AI 기업 네이버의 뉴스 저작권 가치 폄훼 유감" (November 6, 2026)
- iNews24: 한국방송협회 "네이버, 뉴스 저작권 가치 폄훼 유감" (November 5, 2026)
- MLex: Naver faces compensation claim from South Korean broadcasters over AI training (Nov. 5, 2026)
- MLex: South Korea's Naver sued by local broadcasters over training of AI models (Jan. 14, 2025)
