<\/p>\n
South Korea’s publishing sector is advocating for clear labelling rules to distinguish AI-assisted books from human-authored works, as debates about transparency and quality intensify globally.<\/p>\n<\/div>\n
South Korea\u2019s publishing industry is moving towards clearer rules on artificial intelligence after a growing backlash against so-called \u201cclick-and-publish\u201d books, with industry voices arguing that readers should be told exactly how much machine assistance went into a title. At an emergency forum in Seoul on 29 April, the Korean Publishers Association said the sector needed a visible labelling system to distinguish books shaped by human authorship from those heavily produced with generative AI.<\/p>\n
Speakers at the event warned that existing safeguards are no longer enough. Professor Park Jeong-in of Duksung Women\u2019s University said current rules could easily allow traditionally edited books to sit alongside works that were largely generated by AI, making it difficult for readers to know who was accountable for the final text. She argued that disclosure should cover which stages of production involved AI, including writing, translation and proofreading, as well as identifying the responsible editor.<\/p>\n
A three-tier classification was also put forward by Yoon Seong-hoon, head of the publishing house Clayhouse. Under his proposal, books would be marked as purely human-made works, AI-assisted works that had been checked and controlled by people, or titles created mainly by AI. He said the information should be shown prominently in places such as the copyright page or the front flap, so that transparency becomes part of the publishing standard rather than an afterthought.<\/p>\n
Concern is being fuelled by the scale of low-quality material entering the system. According to the National Library of Korea, 11,651 e-books were rejected during legal deposit checks in 2024 and 2025, a sign that large volumes of minimally edited content are already placing pressure on public cataloguing and knowledge-management systems. The broader debate is not confined to South Korea: recent industry coverage in the US and elsewhere has shown publishers, authors and platform operators wrestling with disclosure rules as readers demand authenticity and fear that opaque AI use could erode trust. The consensus emerging from Seoul was that the answer is not to ban AI outright, but to make provenance part of the value of a book.<\/p>\n
Inspired by headline at:<\/strong> [1]<\/a><\/sup><\/p>\n