Context: An increasing number of media outlets have been suing AI service providers over copyright infringement across the world. In December 2023, the Korean Broadcasting Association (KBA) became aware of this and warned a handful of tech giants, including Naver, Kakao, Google Korea, and Microsoft against the unauthorized use of their content for AI training. In separate letters to each of the companies, it noted that all forms of video and audio content are also copyrighted works and therefore use of that content should also be compensated.
What’s new: Three members of the KBA – Korean Broadcasting System (KBS), Munhwa Broadcasting Corporation (MBC) and Seoul Broadcasting System (SBS) – yesterday filed a copyright infringement suit against Naver in the Seoul Central District Court (January 13, 2025 Korean Broadcast Association press release). The companies allege that it has been using their content without permission to help train its AI platforms Hyperclova and Hyperclova X, infringing on their copyright, as well as the country’s Monopoly Regulation and Fair Trade Act. They are seeking an injunction and damages (the amount of which remains confidential), claiming that Naver has been asked for compensation for their work twice but has “failed to provide a clear response”.
Direct impact: This may be the first suit launched by media platforms against a major tech company in South Korea but it probably won’t be the last. The KBA told ai fray today that its members “intend to file lawsuits against all big tech companies that have used terrestrial broadcasters’ news content without proper authorization”. While the names of those companies cannot yet be disclosed, the warning letters KBA issued in December 2023 were sent to Kakao, Google Korea, and Microsoft.
Wider ramifications: The first-of-its-kind lawsuit will help South Korea establish some legal standards for the use of data for AI training, which has already started occurring in places like the U.S. but which South Korea really needs, the KBA said in a statement yesterday. AI governance has certainly floated to the top of South Korea’s agenda but this lawsuit could be key in the country establishing itself on the global AI governance stage.
KBS, MBC and SBS are South Korea’s top three broadcasters. The KBA is an organization with 39 domestic broadcast members, including KBS, MBC and SBS. The association represents its members in policy cooperation, business promotion, and committee operations. The KBA also has an AI Taskforce (made up of a mix of the members’s legal/IP personnel) which is in charge of negotiations with AI companies regarding the use of content, as well as communicating with the government about AI-related regulation.
The three broadcasters filed their suit in the Seoul Central District Court under case number 2025가합5105. They are mainly seeking:
- Compensation in the form of damages for Naver’s past usage of their content, including the revenue generated from ads placed on pages that link to their articles.
- An injunction stopping Naver from using their copyrighted work unlawfully.
- A restructuring of how Naver displays their content moving forward, including by crediting the original sources or limiting the ways it is used.
The KBA noted that it approached Naver twice about paying the appropriate compensation for its content use. It asked Naver to be transparent about where it was getting its data to help train its AI software, but the tech company refused, stating:
“The type and details of learning data are difficult to disclose as they are technical know-how that is an important asset of the company.”
That software was then used to operate various services based on its HyperclovaX platform, and gained economic benefits from it, the association alleges. After sending Naver a warning letter in December 2023, the tech company claimed it had already stopped using news content for AI training six months prior, but the broadcasters allege that Naver’s conversational AI service “ClovaX” is still using their news content.
In a statement on its website, the KBA said:
“This lawsuit goes beyond merely protecting the rights of the plaintiffs; it represents a critical turning point in our country for safeguarding the independence and credibility of journalism. It also aims to establish legal and ethical standards for the emerging technology of generative AI.”
Meanwhile, a KBA AI Taskforce official said:
“We will continue to do our best to compensate broadcasters for the use of news data and prevent future infringements. Starting with Naver, we will take strong action by sequentially filing lawsuits against big tech companies that have used news data without permission.”
The KBA has confirmed to ai fray today that the broadcasters intend to sue all companies that have used their content without proper compensation. Those companies remain unnamed for now.
The association also noted that this lawsuit will be an “opportunity” for South Korea to set clear guidelines for the use of data in AI training.
The country has already made several moves to establish itself on the global AI governance stage, including hosting the AI Seoul Summit 2024, where governments from some of the world’s largest economies signed the Seoul Declaration for Safe, Innovative and Inclusive AI (May 23, 2024 Ministry of Foreign Affairs press release). It also passed its own legislation criminalising sexually explicit deepfakes (September 26, 2024 Reuters article), enacted its AI Basic Act (December 27, 2024 Korea.net article), established an AI Safety Institute (November 27, 2024 Korea Times article) and issued a GenAI competition report proposing merger control regime amendments (December 20, 2024 ai fray article).The suit also comes after AI service providers across the globe have become targets in copyright infringement suits launched by media companies – OpenAI being a principal one. The ChatGPT-owner is embroiled in dozens of disputes in the U.S., including by the New York Times (June 12, 2024 ai fray article), as well as in Canada (December 3, 2024 ai fray article) and India (November 18, 2024 ai fray article). Most of those suits claim that OpenAI has been using their content to train its software too.