Courts in Brazil and Norway block Meta’s bid to exploit news content without permission

A Brazilian court has stopped Meta in its tracks. On 3 November, a São Paulo court temporarily suspended the disputed provisions of the company's new Commercial Terms, which had been due to take effect on 20 November. Acting on a complaint filed by Brazilian associations representing newspapers (ANJ), magazines (ANER), and radio and television broadcasters (ABERT), the court prohibited Meta from collecting and using their members' content without authorisation. The ruling also stressed the need to obtain publishers' explicit consent before using their content to train AI models, according to an ANJ statement.

The original version of the terms would have authorised Meta to collect content to develop, improve and "train" its products - a term associated with the development of AI models. This could have included articles, photographs and videos published on media outlets' websites, not just the content they published on Facebook or Instagram. The court also ordered Meta to maintain the accounts and tools of media outlets that rejected the terms, without taking retaliatory action against them. Any breach of the order would expose Meta to a daily fine of 1 million Brazilian reais (around 163,000 euros), capped at a total of 100 million reais.

The ruling follows a decision by the Oslo District Court in Norway on 28 October. Ten Norwegian media companies and groups had challenged the same terms, which would have forced them either to allow Meta to collect their content, including paywalled material, or to leave its platforms. There too, the court put a temporary stop to Meta's plans, pending a final ruling on the case.

In the United Kingdom, the Competition and Markets Authority (CMA) announced on 30 October that it was opening an investigation into the new terms following a complaint from representatives of news publishers. The regulator expressed concern that the terms could unlawfully impose "unfair trading conditions", amounting to an abuse of a dominant position. The European Commission has also confirmed that it has received a complaint on the matter.

Meta backs down, but uncertainty remains

The new Commercial Terms were initially due to take effect on 30 October 2026, prompting fierce opposition from news publishers. Following pressure from the media industry and the public, Meta removed the word "train" from the terms and postponed their implementation until 20 November. The revised version nonetheless retains broad permission to collect content for Meta's advertising and commercial products, as well as to maintain their integrity and security. On 28 October, Meta announced consultations with industry representatives and postponed implementation of the terms indefinitely.

A stark imbalance of power

Meta's attempt highlights a worrying trend: major technology companies are seeking to expand their access to news content without guaranteeing fair and sustainable payment to media outlets. Journalistic content is being used to train the AI models that power conversational AI tools, which are increasingly acting as gateways to information. These tools are therefore competing with the very news organisations whose content they exploit.

Partnerships between news publishers and AI developers are one solution being considered by the industry, but they remain woefully inadequate and reflect stark geographical inequalities. According to an RSF investigation, 71% of these agreements involve only Western English-language media outlets.