South Korea Launches AI-Copyright System Improvement Working Group

The Ministry of Culture, Sports and Tourism of South Korea (MCST), in collaboration with the Korean Copyright Commission, announced the establishment of the "2024 AI-Copyright System Improvement Working Group." The initiative aims to proactively address copyright issues emerging in the era of artificial intelligence (AI) and to develop concrete policy alternatives for key discussion points.

Following the operation of the "2023 AI-Copyright System Improvement Working Group," the MCST published a "Generative AI-Copyright Guide" in December last year. This guide outlines considerations for AI operators, users, and rights holders regarding the use of generative AI and the copyright registration of AI-generated outputs.

This year, the working group will focus on ongoing issues, including the manner in which copyrighted works can be utilized for AI training, the disclosure of training data, and the legal nature and copyright infringement issues concerning AI-generated outputs. The group comprises experts from academia, legal circles, rights holders, business operators, and the technology industry. It aims to provide a comprehensive and multi-faceted discussion on these topics, ensuring diverse and objective perspectives by including recommendations from related government departments, courts, and prosecutors from the outset.

A survey by the Ministry of Science and ICT in December 2023 highlighted concerns in the digital era, with the scope of use of AI-generated works and the recognition of copyrights for AI creations ranking as top issues among respondents.

The MCST plans to conduct the working group through monthly meetings and subcommittee discussions from February to November. These sessions will cover specific issues, including legal use of copyrighted materials for AI training and copyright registration criteria for AI outputs. The meetings are designed to foster detailed policy discussions and share findings for further refinement.

The relationship between AI and copyrighted works is examined through two phases: the collection and use of data, including copyrighted works during the training phase, and the creation of data in the output phase. This year, the discussions are divided into two subcommittees focusing on these respective areas, covering topics such as lawful use of copyrighted works for AI training, disclosure of training data, copyright protection for AI outputs, and criteria for copyright infringement by AI-generated outputs.

In parallel with the working group's operations, the MCST and the Copyright Commission will conduct a study on "AI-Copyright Law and System Improvement." The results will inform comprehensive strategies to address AI-copyright issues, with plans to gather public feedback and announce policy directions by the end of the year.

Globally, nations are recognizing the imperative to adapt copyright laws for the AI era. For instance, the European Union recently reached a consensus on the final version of its AI Act, aiming to regulate AI applications while ensuring innovation and ethical standards. Similarly, the United States has issued executive orders directing the establishment of frameworks to manage AI technologies and their impact on copyright. These cases highlight a worldwide effort to balance the advancement of AI technologies with the protection of intellectual property rights, ensuring fair use and fostering innovation within the digital economy.