Danish publishers sue OpenAI after failing to strike rights deal

Published connected 01/07/2025 - 13:28 GMT+2•Updated 13:29

An manufacture group representing almost each of Denmark’s media outlets including broadcasters and newspapers has said it's suing ChatGPT’s genitor institution OpenAI for utilizing its content, successful a statement issued connected Tuesday.

“Big tech must respect nationalist laws, and arsenic such, DPCMO has a work to act. Pursuing ineligible action is not a choice; it’s a necessity,” nan Danish Press Publications’ Collective Management Organization (DPCMO) said, claiming that: “OpenAI has utilized DPCMO repertoire successful its training, [and] OpenAI does not respect reservations successful position & conditions of a website.”

DPCMO said it reached retired to OpenAI “long ago”, successful a bid to study astir nan product, data, exertion and nan institution arsenic they usage accusation from DPCMO’s members “to train its app ChatGPT”. 

But an effort to negotiate pinch nan publishers arsenic a corporate failed, according to nan group, and OpenAI told it "they would not prioritise Denmark”, DPCMO said.

Nor did an involution by nan country’s Minister for Culture succeed, nan group claimed.

A spokesperson for OpenAI told Euronews that nan institution "already held nonstop talks pinch nan DPCMO connected this rumor and arsenic a consequence we do not usage its members’ contented for hunt results aliases training. We person not received immoderate accusation astir ineligible proceedings but are assured successful our position.”

Until now, OpenAI has been striking deals pinch publishers individually, specified arsenic pinch British newspaper nan Guardian, and France’s Le Monde.

The institution agreed successful June 2024 to region property publications of DPCMO's members from training sets.

Code of practise

The action - revenge astatine nan City Court successful Copenhagen - comes arsenic Denmark starts chairing the meetings of EU ministers connected Tuesday.

AI and copyright are astatine nan forefront of ongoing discussions betwixt nan European Commission and rightsholders arsenic nan EU executive is group to publish the Code of Practice connected general-purpose AI (GPAI) earlier 2 August.

The Code, a voluntary group of rules, intends to thief providers of AI models, specified arsenic ChatGPT and Gemini, comply pinch nan EU’s AI Act. The last type of nan Code was group to travel retired early May but has been delayed.

The process has been criticised throughout, by tech giants arsenic good arsenic publishers and rights-holders concerned that nan rules break nan EU’s Copyright laws.

The usage of imaginative contented by AI devices and companies has sparked debates, not only successful Europe, but globally. In 2023 Hollywood writers fought to protect imaginative activity from nan usage of ample connection models for illustration ChatGPT.