KBS headquarters in Yeouido, Seoul / Korea Times file
Korea's three terrestrial broadcasters filed a copyright lawsuit against Naver, Tuesday, accusing the major Korean online platform of using their news content for generative artificial intelligence (AI) training without permission, according to the Korean Broadcasters Association.
In the lawsuit, KBS, MBC, and SBS claimed that Naver used the broadcasters' news articles for its AI platforms, HyperCLOVA and HyperCLOVA X, without permission, thereby infringing on their copyrights and violating the Unfair Competition Prevention Act.
In December 2023, the association sent notices to both domestic and foreign information technology (IT) companies, including Naver, Kakao, and Google Korea stating that "separate compensation consultations are required to use news content for AI learning” and that they are prohibited from using news content without obtaining permission first.
Additionally, the association, which represents 39 broadcasters, established an AI task force composed of legal and intellectual property experts.
The task force made two formal requests to Naver for compensation for the use of news data from broadcasters and for measures to prevent further infringement, but received no clear response, according to the association.
"In response to our request for disclosure of the source and content of AI learning data and the data acquisition path, Naver refused to do so," the association's task force said. "Starting with Naver, we will file lawsuits against other big tech companies that use news data without permission and take a strong stance."
A Naver official said, "The lawsuit filed by the Korean Broadcasters Association has been confirmed through news reports, but we have not yet learned the contents of the suit. We will be able to clarify our position after understanding the details."
Meanwhile, Naver has previously defended its position, asserting the fair use of news content for training generative AI.
However, following the emergence of the issue, the company has refrained from using news data without consent, after revising its terms and conditions in response to protests from media outlets.
Naver plans to engage in public discussions and consider media feedback on the issue.
Naver CEO Choi Soo-yeon delivers a speech about the company's hyperscale artificial intelligence platform HyperCLOVA X during a press conference in Seoul, Aug. 24, 2023. Yonhap
Meanwhile, the Korean Association of Newspapers has also been reviewing the issue with legal experts to prevent any copyright infringements.
“Although the association has not yet decided to file a lawsuit like the Korean Broadcasters Association, we are considering various measures regarding IT companies' violation (of news content copyrights),” the association's official said.
On Dec. 28, 2023, the newspaper association submitted a request to the Fair Trade Commission, urging the correction of Naver's unfair use of news content for generative AI training.
The association argued that Naver's use of news content for AI training is an act without prior consent from the media and constitutes copyright infringement.
It also called on the government to specify relevant provisions in the partnership's terms and conditions, ensuring that media companies receive proper compensation for the use of news content in training HyperCLOVA X.
Regarding Naver's refusal to disclose the size and scope of news data used for Generative AI learning to media companies, the newspaper association said, "The government should force Naver to disclose the information and purpose of use of news data, which is the basis for calculating the news usage fee used for HyperCLOVA X's learning."