The European Publishers Council has written a second time to the European Commission about Meta's revised Commercial Terms and Self-Serve Ad Terms, asking Brussels to examine them under the Digital Markets Act and EU competition law and, if it cannot adopt interim measures before November 20, 2026, to obtain an immediate undertaking from Meta to postpone the contested provisions. The Brussels-based trade body for news publishers published its statement on October 23, 2026, the same day Meta moved the effective date of the terms from October 30 to November 20. Its verdict on the revisions Meta has made so far is blunt: "little has changed in substance."

The statement is an advocacy position from a party to the dispute, not a regulatory finding. The council has not published the letter itself, and the statement does not say when it was sent or whether the Commission has replied. Angela Mills Wade, the council's executive director, promoted it on LinkedIn the following morning, on October 24.

In Short

A group representing big European news publishers has complained to the EU for a second time about Meta's new rules for businesses on Facebook and Instagram. The publishers say the rules still let Meta's robots collect content from their websites for uses that have nothing to do with running a page or buying an ad, and that there is no way to say no while staying on the platforms. They want the EU to check whether the rules break competition law and to stop the disputed parts from starting on November 20 until that check is done.

What the council says has not changed

The core of the statement deals with the one change Meta has made to the disputed clause. "Meta has removed the express reference to 'train' from its website-scraping provision and attached its stated purposes around advertising and business products, alongside the integrity and security of Meta's products," according to the council.

That matches the revised text. When PPC Land first examined the incoming terms on October 16, the new Section 1 let Meta crawl business websites and use their content to train AI from October 30. A day after The Verge reported the training sentence had been dropped, a check of the UK English preview still found it in place. By the evening of October 23, the preview had caught up: the free-standing training sentence was gone, and the crawling authorisation ended instead with a purpose limited on its face to providing, developing, improving, researching and maintaining advertising and business products, and to maintaining the integrity and security of Meta's products.

The council does not accept that this settles anything. "However, removing that one word does not establish an express exclusion of AI training, fine-tuning, grounding, or the generation and display of answers and summaries," the statement reads.

The distinction is legal rather than semantic. A contract that lists permitted purposes does not necessarily forbid others unless it says so, and "advertising and business products" is not defined in the preview. The News/Media Alliance's open letter to Mark Zuckerberg made a similar point on October 22. What the European council adds is a list of specific AI uses it wants written out of the contract, not just a word taken out. Grounding, in which a model retrieves live web content to answer a query, and the display of generated answers and summaries are uses that do not require training at all. A clause that only removed "train" would, on the council's reading, still leave room for them.

Mills Wade put the same point more directly in her LinkedIn post: "Meta has removed the explicit reference to AI training but its terms still seek sweeping permissions to crawl, collect, analyse and reuse content from businesses' websites. Crucially, they do not expressly exclude AI grounding or the generation of AI answers and summaries."

The scope of the authorisation

The statement also lists the activities the clause still covers. "On the contrary - the authorisation remains extensive," according to the council. "It covers automated access, indexing, caching, analysis, collection and retrieval of website content, the creation of test credentials and the periodic collection of updates."

Each item corresponds to wording in the preview. Businesses authorise Meta, or any entity acting on its behalf, to use web scrapers, bots, robots, spiders, crawlers and user agents, to create test credentials, and to access, index, cache, analyse, collect, retrieve or crawl domains associated with their products or services, pulling updates from them periodically. The test credentials provision, which PPC Land flagged in its first analysis because the terms do not explain what such credentials are, has survived both revisions.

The council asks Meta for an explanation of three things. "Meta needs to explain precisely what it intends to collect, why that collection is necessary and how the material may subsequently be used," according to the statement.

For publishers the concern centres on a single trigger. "Maintaining a business Page or advertising a subscription should not confer general rights over a publisher's website and archive," the council writes. "Information promoting a publication must be distinguished from the journalism within it." Because the clause reaches any URL associated with a business's products or services, a newspaper that advertises a subscription offer on Instagram could, on that reading, be authorising access to its whole site.

No mechanism to refuse

The second strand of the complaint deals with consent. "The proposed terms provide no express contractual mechanism allowing publishers to refuse the additional permissions while continuing to use Meta's business services," according to the council.

Section 7 of the preview explains why. Continuing to use Meta products for business purposes after notice of an update means agreeing to it; the only alternative offered is to stop all business or commercial use of Meta's products. There is no way to accept the terms for a Facebook Page and refuse them for a website. PPC Land described this all-or-nothing structure in its October 17 piece on how Meta grants itself the right to crawl advertisers' websites, noting that agreement is implied by continued use and there is no partial opt-out.

The council wants that structure reversed. "Uses beyond those necessary to provide the specific service requested should require separate, express authorisation from the relevant rightsholder," the statement says, and publishers "must remain free to negotiate the scope, value and price of additional uses, or refuse them, without losing access to those services or suffering penalties."

That position echoes European case law that publishers have already used against Meta. In May, the EU Court of Justice backed Italy's publisher remuneration rules in a Meta challenge, on the condition that publishers keep the right to refuse authorisation or to grant it free. In July, France's competition authority ordered Meta to return to good-faith talks with publishers within 15 days through interim measures. Both cases turned on the publisher's ability to set terms, which is what the council says the Commercial Terms remove.

Why robots.txt is not enough, in the council's view

Meta's main public answer to the criticism has been that site owners can still control access through robots.txt, the file at a site's root that tells crawlers which paths they may fetch. The council rejects that as a substitute for contract language. "Technical controls over crawler access cannot substitute for clear, enforceable contractual limits on collection and subsequent reuse," according to the statement.

There are two parts to that objection. The first is legal: a robots.txt directive is a request that a crawler may or may not honour, while the business authorisation in the terms is a grant that the business has signed. If the two conflict, it is not clear from the preview text which prevails, because the terms do not mention robots.txt at all.

The second part is about identification. "Publishers also need to know which automated agent is acting, on whose behalf and for what purpose," the council writes. Meta runs several crawlers. Its Meta-ExternalAgent user agent collects public pages to train AI models and index content for Meta products, which makes it a mixed-use crawler whose single name covers more than one purpose. A publisher that blocks it to keep content out of AI systems may also block uses it wants to allow, and the clause extends to any entity acting on Meta's behalf, without naming any agent. The same demand, that bots disclose who they are and why they are crawling, sits behind the Stealth Bot Prohibition Act that publishers have lobbied Congress to pass in the United States.

Liability on one side

The statement also ties the crawling permissions to the rest of the contract. "The permissions must be assessed alongside existing warranty, indemnity and liability provisions that place substantial responsibilities on businesses while sharply limiting Meta's own contractual liability," according to the council.

The preview's indemnity clause was widened in this revision to cover claims brought by the business itself, PPC Land found in October. The new AI Features section goes further, making advertisers solely responsible for edits Meta's AI tools make to their ads, including features Meta turns on. The council does not cite specific section numbers, but its argument is that a broad grant of rights to Meta should be read together with a contract that leaves the business carrying most of the legal risk.

What the council wants the Commission to do

The request to Brussels has two layers. "We are asking the Commission to scrutinise the proposed conditions under the Digital Markets Act and EU competition law, including whether access to designated services is being made conditional on unfair terms," according to the statement.

The reference to "designated services" points at Meta's status as a gatekeeper. Facebook and Instagram are among the Meta services designated under the Digital Markets Act, although the General Court in June annulled the designation of Facebook Marketplace while upholding Messenger's. Article 6(12) of the regulation requires gatekeepers to apply fair, reasonable and non-discriminatory general conditions of access to business users of their online social networking services, among others. The council does not name the article, but the wording of its request maps onto it.

The second layer concerns timing. "If interim measures cannot be adopted before 20 November, we've asked the EC to seek an immediate undertaking from Meta to postpone implementation of the contested provisions until the Commission has completed its assessment and determined what action is required," according to the statement.

That is a narrower demand than a full suspension. The council is not asking Meta to withdraw the whole Commercial Terms, which also cover routine matters. It wants the contested provisions held back until the Commission has finished its assessment. Its argument for urgency is irreversibility. "Once content has been collected and incorporated into datasets, models or competing services, restoring control may be extremely difficult," the statement reads. Interim measures exist in EU law for precisely this kind of claim of serious and irreparable harm, and publishers have asked for them before: an antitrust complaint against Google's AI Overviews, filed with the Commission in June 2025 by the Independent Publishers Alliance, sought interim measures on the same grounds.

How fast the Commission can move is another question. Interim measures under competition law require the opening of proceedings and a prima facie finding of an infringement. A three-week window is short for either. Meta's own explanation for the delay, given to Reuters on October 23, was that it needed more time to explain to partners what is and is not changing, not that it would rewrite anything.

Beyond publishers

The statement presents the dispute as one that affects every business on Meta's platforms. "These concerns extend across sectors," according to the council. "Businesses use Meta's platforms to distribute content, reach customers and promote their products and services. Access to those functions must not become a means of obtaining permissions beyond what is necessary to provide the service requested."

Mills Wade went further on LinkedIn. "No business should have to surrender rights unrelated to the service it has requested simply to access a platform on which it depends," she wrote, adding that "market power must not become a means of obtaining permissions that could never be secured through fair negotiation."

The framing matters because the Commercial Terms do not distinguish between a newspaper and a retailer. Every brand that runs a Facebook Page, uploads a product catalogue or buys ads through Ads Manager accepts the same contract, and the crawling clause applies to their domains in the same way. For most advertisers, Meta pulling product pages to build ads is expected. The question the council raises is whether the same authorisation reaches everything else on a site, from help pages and research to editorial content, and whether any business can limit it without leaving the platform.

The statement draws a line between remuneration and control. "Copyright is about control as well as remuneration," the council writes. "Publishers and other businesses must remain free to decide how their content is used." Mills Wade's post put the same idea in wider terms: copyright, she wrote, "is fundamentally about control, including the right to decide who may use content, for what purpose, on what terms, and whether to authorise that use at all."

Why this matters for marketers

The marketing community has tended to treat the Commercial Terms dispute as a publishers' fight. The council's statement, and the European pressure behind it, makes that harder. Its central demand, a contractual right to refuse uses beyond the service a business requested, would apply to advertisers as much as to newspapers if regulators adopted it.

The calendar is now tight. The News/Media Alliance asked Meta for a public response by October 26. The new effective date is November 20, and acceptance under Section 7 happens automatically for any business that keeps using Meta products after that date. A Commission decision on interim measures, or an undertaking from Meta, would have to arrive within four weeks of the council's letter to change that outcome.

Reaction under Mills Wade's post came from the publishing side. Matthias Stöcher, whose profile lists him in legal and policy at the Austrian newspaper DER STANDARD, thanked her "for the quick and detailed response to this outrageous behavior." Neither Meta nor the Commission is quoted in the council's statement. The statement does not say whether other publisher groups co-signed the second letter, and it does not include any figures on how many council members use Meta's business products or what share of their traffic comes from Facebook and Instagram.

What it does establish is that removing one word has not ended the European objection. Meta now faces publisher trade bodies on both sides of the Atlantic asking for more than clarification, and a date that, unless it moves again, will make the current text binding on every business that stays.

Timeline

Summary

Who: The European Publishers Council, a Brussels-based trade body for news publishers, and its executive director Angela Mills Wade; Meta Platforms; the European Commission; publishers and other businesses bound by Meta's Commercial Terms and Self-Serve Ad Terms.

What: The council wrote to the Commission a second time, arguing that Meta's removal of the word "train" from its website-crawling clause does not exclude AI training, fine-tuning, grounding or AI-generated answers and summaries, that the terms give businesses no way to refuse the extra permissions, and that robots.txt cannot replace contractual limits. It asked the Commission to examine the terms under the Digital Markets Act and competition law, and to secure a postponement of the contested provisions if interim measures cannot be adopted in time.

When: The council published its statement on October 23, 2026, and Mills Wade shared it on LinkedIn on October 24. The terms are due to take effect on November 20, 2026.

Where: Brussels, where the council and the Commission are based. The terms apply worldwide to business use of Meta's products.

Why: Under Section 7 of the terms, any business that keeps using Meta's products after November 20 accepts them in full. The council argues that once website content has been collected and built into datasets, models or competing services, publishers cannot get control back, which is why it wants action before the terms apply.