Today the European Parliament adopted the report on “copyright and generative artificial intelligence” drafted by MEP Axel Voss (by 460 votes to 71, 88 abstentions). The European Federation of Journalists (EFJ) welcomes this first step as a sign of political will to strengthen and make fairer the copyright framework protecting journalists from the unauthorised use of their work in AI training and content generation.
Until now, AI companies have been relying on journalistic content without seeking consent, providing remuneration, or ensuring transparency about how such material is used. This practice not only undermines journalists’ intellectual property rights but also threatens the economic sustainability of journalism, while tech companies reap huge profits. And even when agreements providing for remuneration are signed, they do not concern small and diverse media outlets and neither for whole media markets in whole countries. By creating an imbalance of power where only the largest media companies in the largest European markets have access to licensing deals, this poses a serious threat to media pluralism, independent journalism, and diversity of opinion in the news.
In this context, with the 2019 Copyright Directive in the Digital Single Market (CDSM) due for review in 2026, the European Parliament is calling for swift solutions to protect right holders and determine whether existing copyright rules cover AI training.
The report is an important first step in the right direction for journalists as it calls to give authors control over the use of their content by AI while reinforcing transparency and accountability obligations for AI developers. It promotes fair and proportionate remuneration through licensing and highlights the flaws of the current opt-out model, which is not only impractical and lacks transparency, but also in contrast to European copyright practice. Finally, it tries to integrate AI-related concerns into existing EU legislative frameworks on media freedom, pluralism, and democratic discourse.
In particular, the report acknowledges:
- The limitations of the Directive on Copyright in the Digital Single Market (DSM) in addressing new issues created by generative AI systems;
- That rights may also be managed under a transparent and voluntary licensing system that restores negotiation power to rights holders, without prejudice to individual licensing;
- The need for total control over AI use of journalistic material, including full transparency on data used from AI developers;
- That the journalists which own the rights to content (articles, photos, broadcasts, etc.) should be able to decide how that content is used, not only for training AI, but also for other AI uses, e.g. using the content to produce answers or summaries;
- The risk AI poses to news diversity and the need to enforce EU law to mitigate this risk with the principles of the EMFA and DSA;
- The need for an effective and transparent opt-out mechanism to allow rights holders to exclude their works from AI training datasets; MEPs stand against the idea of a centralised registry of opt-outs under the Text and Data Mining (TDM) exception of the Copyright Directive.
In addition, the report urges the European Commission to assess and propose mechanisms to ensure that GenAI providers fairly, proportionately, and non-discriminatorily compensate media outlets for diverting traffic and revenue, with particular attention to local and regional media.
“The EFJ welcomes this report as a first step in the right direction and calls on the Commission to act swiftly. The use of journalistic content by AI-Systems should be subject to clear prior authorisation and fair remuneration, protecting moral rights, managed by collective management organisations,” said EFJ President Maja Sever.
Source: EFJ
