MON, OCTOBER 19, 2026
Independent · In‑Depth · Practitioner‑Tested
✎ News

Korean Publishers Tell AI Companies: No Books, No AI

The Korean Publishers Society's national appeal had 158 organisations and 321 individuals signed as of 16 October. It demands AI firms record what they trained on and the legal basis for it, insists silence is not consent, and follows the $1.5 billion Bartz v. Anthropic settlement that worked out to roughly $3,000 per work.

By AIToolsRecap October 19, 2026 6 min read 33 views
Home › Articles › News › Anthropic › 158 Korean Groups Demand AI Firms Disclose Trai...

The Korean Publishers Society has drafted a national appeal titled No Books, No AI. As of 16 October it carried 158 organisations and 321 individuals as signatories, and it is to be delivered to the Korean government, the National Assembly and AI companies directly.

The demands are specific rather than rhetorical. AI companies should record which works they trained on, where those works came from, and the legal basis for using them - and rights holders should be able to verify that record. Silence from a rights holder should not be treated as consent. Small presses and individual creators should be able to negotiate on fair terms rather than accept a standard offer. And the appeal asks the government not to curtail creators' rights in the name of industry growth.

The context is a settlement that set a number. In Bartz v. Anthropic, a US federal judge granted final approval on 20 July 2026 to a $1.5 billion settlement covering roughly 482,000 works - about $3,000 each - over roughly four million pirated digital books downloaded in 2021 and 2022. Korean reporting has also pointed at Nippon Shuppan Hanbai, which began supplying Anthropic in December 2024 for roughly a year, and at domestic distributors Woongjin Booxen, Kyobo Book Centre and Yes24. The appeal follows the Frankfurt Declaration issued on 6 October.

Tags
AI NewsAnthropic2026
⚑

Spot an inaccuracy?

We verify facts before publishing and correct errors promptly. If something in this article is wrong or outdated, let us know.

Report an error →