What is it about? Anyone using Facebook or Instagram for business purposes, that is, running a page or placing ads, was in future to allow Meta to automatically crawl, analyse and cache the websites linked to their account, including the content accessible through them. For a publisher, this would mean: along with the brand's Facebook presence, its own website and archive might also be made available. The clause also permits Meta to set up test accounts. What that could mean for content behind a paywall needs no explanation in this industry.

The original version even explicitly mentioned the training of AI systems. After the first protests, Meta deleted this wording and postponed the launch to 20 November. The introduction has since been suspended altogether until talks with industry associations have taken place. Meta rejects the accusations: the rules, it says, only apply to advertising and business products, and instructions in robots.txt will continue to be respected. This commitment does not appear in the contract text. As long as that remains the case, it is a declaration of intent that no publisher can rely on.

One could consider the uproar exaggerated. Large parts of our websites are publicly accessible and have been captured for years by search engines, aggregators and other systems. Publishers themselves benefit from their content being found and distributed via external platforms. A company adapting its terms of service to new products is not, in itself, a scandal.

The problem is the way consent is obtained. The new terms apply upon continued business use of Meta's services; no separate opt-out is provided. For many publishers, Facebook and Instagram remain important channels for reach and monetisation. Anyone who does not want to accept the clause would, if in doubt, have to give up their business presence there. The Norwegian media houses that have taken Meta to court call this an all-or-nothing decision. It has little to do with a negotiation between business partners.

That Meta may process the content we ourselves publish on Facebook should go without saying. Deriving from this a right of access to our websites and archives goes much too far for me. The deleted AI passage changes little, because from the publishers' point of view it remains unclear for what purposes the captured data may ultimately be used.

Economically, the matter is particularly annoying. For months, the industry has been wrestling with the question of how media companies can be remunerated for the use of their content by AI companies, while AI-generated answers are pushing down click-through rates. Meta itself has concluded paid agreements with individual media companies, thereby acknowledging that journalistic content has economic value. Why, then, should the use of comparable content from other publishers be governed by the terms and conditions of an advertising account?

What I find remarkable is the reaction of the media industry. Within a few days, WAN-IFRA, the European Publishers Council and numerous national associations spoke out. In Germany, the newspaper publishers' association BDZV and the magazine publishers' association MVFP called on the EU Commission to examine the clause and prohibit it for the time being. In Norway, ten media houses, including NRK and Schibsted, obtained a preliminary injunction against the new terms before the Oslo District Court. With the CMA investigation, the legal question, which often remains abstract in such debates, has now become very concrete.

Meta presumably did not put its terms on hold out of pure insight. The joint pressure has had an effect. Whether the company is prepared to make substantial changes will be shown by the announced talks. A further postponement or a linguistic smoothing of the clause is not enough. The industry needs binding commitments in the contract text, on robots.txt as well as on the purpose for which the captured content is used, and a separate licence for any further use. Above all, it must be possible to refuse the use of one's own content without giving up one's business presence on Facebook or Instagram.

Until now, media houses have too often negotiated individually with the platforms, while the platforms set the technical and contractual framework largely on their own. The past few weeks have shown how much weight the industry carries when it acts united internationally. We should maintain this unity, even if the next clause makes fewer headlines. Meta wants our content. Then Meta should negotiate for it.