The UK is investigating Meta over its publisher terms: web content becomes an AI battleground
Direct answer
On October 30, 2026, the UK Competition and Markets Authority (CMA) opened a formal investigation into Meta's commercial terms for businesses using Facebook, Instagram and its other services. The investigation's question is whether those terms let Meta collect content from those businesses' websites and use it to develop and provide its own products and services, including AI, without giving them a choice over whether or how their content is used. The CMA acts under Chapter II of the Competition Act 1998, which prohibits abuse of a dominant position. No decision has been made on whether Meta broke the law.
The timeline: a pause did not stop the investigation
- The terms were due on October 30: the terms revision was scheduled for that date. After industry complaints, Meta moved the effective date to November 20.
- On October 28, a pause and a Norwegian ruling: Meta announced it was pausing the rollout while consulting industry groups. That same day, an Oslo court temporarily blocked the terms, in a case from the Norwegian Media Businesses' Association for ten media groups, including NRK, TV 2 and Schibsted.
- On October 30, the CMA opened the case anyway: the investigation also covers the existing terms, not only the paused version. That wording matters: the contract businesses accept today falls within scope, not only the text on hold.
- Who drove the case: the CMA acts on concerns from the UK News Media Association and the Movement for an Open Web, which filed complaints with the CMA and the European Commission on October 22. The first round of evidence gathering is scheduled to run until January 2027, per coverage.
- Meta's response: the company said there are no grounds for an investigation and that it already committed not to change its terms without consulting publishers and other stakeholders.
What it means for your business
- If you publish on Meta platforms, your terms matter: the investigation concerns the terms a business accepts when using Facebook and Instagram. Review which version you accepted and what usage rights over your site's content it contemplates, with legal advice if content is a core asset of your business.
- Two different fronts in the same week: on the same October 30, the four big AI labs backed a shared token for expressing a no-training preference in robots.txt. That token governs model training; Meta's terms govern the commercial contract. They are different tools and should be treated separately.
- The precedent at stake: if the CMA finds abuse, the remedy under discussion is opt-in licensing, and that standard would reach any lab training on open web content. For now, the investigation is an open question, not a ruling.
- The practical lesson: your site's content has value as an AI input, and that value is negotiated in three places at once: your robots.txt, your terms with each platform, and each country's regulation. Reviewing only one of the three leaves the other two open.
Want AI that works for your business?
Message us on WhatsApp and let's talk about automation and AI agents for your business.
WhatsApp inicIARelated articles
- The four big AI labs backed one switch to opt out of training
- Privacy and data: what happens to your customers' information
- What an AI agent is and how a real business uses one
Frequently asked questions
What is the CMA investigating about Meta?
Whether Meta's commercial terms for businesses using its platforms amount to abuse of a dominant position, by letting it collect content from those businesses' websites and use it for its own products, including AI, without giving them a real choice.
Has Meta already withdrawn the terms?
No. Meta paused the rollout on October 28 while consulting industry, and an Oslo court blocked them temporarily. The CMA investigation goes ahead and also covers the existing terms.
Has it been decided that Meta broke the law?
No. The investigation opened on October 30, 2026 and the first evidence round is scheduled until January 2027.
Is this related to the no-training token?
They are different fronts. The shared token governs the no-training preference in robots.txt; Meta's terms govern the commercial contract a business accepts when using its platforms.
