Who will pay for news in the AI era: a discussion at the Global Media Forum in Berlin

AI answers with news, but journalism remains invisible

AI chatbots are increasingly becoming a source of news: according to a 2026 Reuters Institute survey covering more than 90,000 people worldwide, 10% of users turn to chatbots for news weekly, and among younger audiences this share is noticeably higher. This was discussed at the Global Media Forum (GMF) in Berlin, where publishers, lawyers and technology companies debated who should pay for the journalism that AI systems use, Deutsche Welle reports.

A broken exchange: the traffic is gone, the content remains

Two forum participants summed up the essence of the problem. Julia Becker, publisher at the Funke media group, recalled that search engines like Google used to index publishers' sites and send readers to them, generating traffic in return — 'but now that exchange has broken down.' DW Director General Barbara Massing noted that AI companies rely on high-quality journalistic material to generate answers, yet the material itself 'often remains practically invisible' by the end of this process. 'AI systems should not simply use quality journalism. They should pay for it,' Massing said.

What is already happening: appeals, lawsuits and courts

Pressure on 'big tech' companies is growing from several directions. In October, German broadcasters ARD and RTL, together with the Federal Association of Digital Publishers and Newspaper Publishers (BDZV) and the Association of Private Media (VAUNET), called for stronger protection of media companies' copyright. In the United States, lawsuits have been added: The New York Times sued Brave Softwar, accusing it of using content without permission and circumventing paywalls. Meanwhile, EU judicial bodies, including the European Court of Justice, are examining whether AI-generated answers themselves may violate copyright. And in February, a coalition of media organizations — the Guardian, the BBC and the Financial Times — demanded rules giving publishers more control over how AI systems copy their content.

A royalty model: paying for every answer

Professor Martin Senftleben, director of the Institute for Information Law at the University of Amsterdam, distinguished two ways AI uses journalistic material: training models and generating answers to specific user questions. Payment for training is a legally more complicated matter; the more urgent problem, in his view, is the real-time use of journalism when generating answers. He believes existing European copyright law already allows publishers and journalists to demand payment collectively, without waiting for new legislation: 'What are we waiting for? Germany is a major member state [of the EU]. This should be introduced in Germany.' The proposed model is mandatory levies for every AI answer that draws on journalism, which a collecting organization would distribute among publishers and authors, modeled on music royalties; the rates of such a system could serve as a starting point in individual negotiations with AI companies.

The industry's position: Google pilots and warnings against regulation

Not all participants in the discussion see the solution in new rules. Lutz Maché, senior manager for government affairs and public policy at Google, said the company is ready to negotiate with German media and discuss new AI products with them. Google is already testing an 'AI Contribution' pilot with individual publishers — it is meant to reward content that substantially influenced answers from products like Gemini and AI Overviews, but this is still an early experiment, not a publicly available payment system. According to Google, its neighboring-rights agreements cover more than 800 German publications, for which payments are already being made. Microsoft, Anthropic and OpenAI recently announced support for standardized signals that allow website owners to indicate whether their content may be used for AI training.

Criticism: a risk for European startups

Joerg Heidrich, a technology law attorney and AI commissioner at the German media company Heise, spoke out against additional regulation. In his view, Europe has already fallen behind in the global AI race, and new rules could make it harder for European companies to compete with major providers from other regions. He also drew attention to smaller players — startups, independent developers, and research institutions that need access to data for training models: the copyright debate, he warned, risks overlooking them, and new restrictions could harm their work. Thus, the forum recorded a split within the industry itself: the media sector insists on payment and control, part of the tech community calls for caution with regulation, while a concrete payout mechanism so far exists only in the form of pilots and proposals.