On 30 October 2026 the UK Competition and Markets Authority opened a Chapter II investigation into Meta's commercial terms. The CMA says the terms could let Meta collect content from the websites of businesses and publishers that use its services. Meta could then use that content to develop its own products, "including its AI features", without those businesses getting a choice on if or how their content is used.
The case is worth backing, and worth keeping narrow. It tests whether a general-purpose competition statute can resolve a dispute over AI data use quickly, without anyone writing a new AI rulebook.
The strongest case for intervention
The argument for acting is serious. For many publishers and advertisers, Meta's platforms are close to unavoidable routes to an audience. If a business must accept whatever terms Meta sets or lose that reach, the "choice" to accept is thin. That is the core of an exploitative-abuse theory. The CMA says the terms "could amount to an unlawful imposition of unfair trading conditions for thousands of UK businesses that rely on its services to reach customers and audiences."
Trade bodies told the regulator the same thing. The CMA names the News Media Association and Movement for an Open Web as having raised concerns. According to The Next Web's reporting, UK publishers argued that Meta has so much influence over their businesses that they cannot realistically refuse its terms. Once content is absorbed into an AI system, the harm may be hard to undo. That is the situation in which an early regulatory look is justified.
What the CMA has and has not done
The CMA has not found a breach. Its case page says "no decision has been made on whether Meta has broken competition law." The investigation must first establish whether Meta holds a dominant position in one or more relevant UK markets. Only then does it ask whether the existing terms, or the revisions to them, amount to an abuse under Chapter II of the Competition Act 1998.
The press release notes that the CMA may use interim measures where temporary action is needed to prevent significant damage. It does not say any have been imposed or sought. The indicative timetable is "rapid initial information gathering" from October 2026 to January 2027, and the CMA says the dates may change. Chief Executive Sarah Cardell said the CMA will work "to reach a swift and satisfactory resolution."
The route matters. The CMA notes that Meta has not been designated with strategic market status under the digital markets regime. It therefore has limited powers under that regime, and the next set of SMS proposals is due before the CMA Board in spring 2027. The CMA is using the older, evidence-led tool instead. That tool requires dominance to be proven, abuse to be demonstrated, and a company to be heard.
Why the sequence of events is the policy lesson
The timeline shows the process working. Euronews reports that Meta had planned to introduce the terms on 30 October. After "active engagement with the CMA", the CMA says, the date moved to 20 November. On 28 October Meta announced a pause to the rollout pending consultation with industry groups. Meta says it has "already committed not to change our terms without consulting publishers and other stakeholders", and says it welcomes the chance to explain its approach to the CMA.
Meta has also disputed the reading of its terms. The Next Web reports a spokesperson saying the changes "were not intended to allow the scraping of news websites." Meta said they were meant to collect product information from shopping websites for advertising, and it removed the word "train" from the wording. That dispute is why a fact-finding process is better than an assumption. A term drafted broadly enough to cover scraping may be a drafting failure, or it may be a real claim on third-party content. The investigation should say which.
A pause that came from regulatory scrutiny, before any penalty, is the outcome a proportionate regime should want. The harm was averted before it began.
Where the risks lie
The pro-innovation concern is not that Meta gets examined. It is what the remedy turns into. AI development depends on access to information. Overbroad rules that treat every automated collection of web content as suspect would hurt startups, researchers and smaller publishers who rely on open access. The sensible line separates two things:
- Content a business supplied to a platform under terms it could not realistically refuse, which is the CMA's actual concern.
- Publicly available content on the open web, which is governed by a separate and unsettled debate.
The CMA should resolve this case on the first of these and should not stretch it to the second.
Three features would make a good outcome:
- A genuine opt-out. Businesses should be able to say whether their content feeds Meta's AI features, without losing access to advertising or distribution tools.
- Plain drafting. Terms should state what is collected and for what purpose. Binding commitments are better than the verbal assurances that publisher groups have called insufficient.
- Speed. A resolution that arrives after the terms have been in force for a year protects no one. The CMA has set a tight initial window and should keep to it.
The CMA also says it is seeking information on Meta's new Meta One business subscription and on Federation of Small Businesses concerns about posting external links. It states that these issues are outside the scope of this investigation. That discipline is right. Folding every grievance into one case would slow it down and blur the legal theory.
The wider signal
Euronews reports that courts in other jurisdictions have already acted, including a Norwegian court that temporarily banned Meta from enforcing the new user terms and a Paris appeals court ruling. Fragmented, court-by-court responses create uncertainty for every company building AI products. A quick, evidence-based UK decision on a defined question would add clarity. If the CMA can show that ordinary competition law handles this dispute with commitments and a short timeline, it weakens the case for heavier, prescriptive AI-specific rules.
The test is straightforward. Success means businesses keep control of their own content, Meta keeps the ability to build products on lawfully obtained data, and the answer arrives within months.