Meta Pauses Publisher AI Terms as Norway Court and UK Probe Escalate

A Norwegian court has temporarily suspended Meta's proposed business terms, and the UK Competition and Markets Authority says it will investigate whether those terms could let the company collect material from business websites without meaningful control. The dispute over Meta AI scraping publishers' websites has moved from complaints into courtrooms and regulators' offices. Meta has also removed the terms from its public page, according to Dataconomy and 404 Media.

The pause itself is not new. Meta first delayed the terms on Oct. 23 and paused them indefinitely on Oct. 28. What changed this week is the legal pressure, and the Nov. 20 date Meta had set for the rollout has passed without a replacement.

Which Meta AI scraping rules are on hold

The proposed terms would have applied to business accounts. They covered four documents: Meta Commercial Terms, Product Catalog Terms, Self-Serve Ad Terms and Ad Creative Generative AI Terms, according to Reuters, reprinted by the Spokesman-Review.

The most contested language let Meta use content from domains, web pages or URLs linked to business accounts to provide, develop, improve, research and train its products. That means a business's linked website could feed Meta's AI systems, alongside the posts it publishes on Facebook or Instagram.

Publishers and digital advertisers complained to EU and UK antitrust authorities. They argued the terms would give Meta unlimited rights to use their content for AI. Meta rejected those claims, and on Oct. 28 it said it was delaying the rollout indefinitely while its existing terms stayed in place.

Those existing terms remain in force while Meta consults. Publishers and business account holders are therefore not operating under the proposed language today.

How the Norway injunction and UK probe raise the stakes

The Norwegian order is a temporary injunction that suspends the terms. The case was backed by the Norwegian Media Businesses' Association and 180 media companies, according to Dataconomy. A temporary order is not a final judgment, and the reporting does not say how long it will last. Still, it turns a publisher complaint into a legal constraint in one European country.

The UK track is a competition inquiry rather than a copyright case. The Competition and Markets Authority said it would examine whether the terms could let Meta collect material from businesses' websites without meaningful control, as MLQ.ai and The 420.in reported. That framing matters, because competition regulators focus on leverage and whether customers can realistically refuse, not only on who owns the content.

For US companies, the bigger signal is the pattern. The EU and UK are testing AI data-access rules that US platforms apply around the world, so a term written for one market can shape practices in others.

California Sen. Adam Schiff wrote to Meta CEO Mark Zuckerberg urging Meta not to implement the terms, as San Jose Inside reported. A letter is not a regulatory action, but it puts the question on the record in Washington.

Timeline from the Oct. 23 delay to the Nov. 20 deadline

The timeline shows a pattern of repeated delays and a document that has now been withdrawn from public view. Each step has pushed the date back or pulled the text down, but none has resolved the underlying disagreement about consent and control.

Why publishers say the terms went too far

Nieman Lab reported that legal experts it consulted concluded the terms would let Meta scrape news sites and use that content to train AI products. That reporting, reprinted by San Jose Inside, is the sharpest legal reading in the coverage reviewed here. It reflects one outlet's consultations, not a court finding.

Publishers argue that unrestricted access weakens their ability to negotiate AI licensing deals. If an AI company can use content without a license, publishers lose the leverage to ask for payment, which sits at the center of the wider fight over AI training data.

Meta sees the matter differently. In an Oct. 28 statement reported by Reuters, a company spokesperson said: "We've heard publishers' feedback on our proposed ads and business terms update." That line acknowledges the criticism without accepting the publishers' description of the terms, and Meta has continued to reject that characterization.

Liability questions and what remains unconfirmed

One secondary report, from Zetik, which summarized Computerworld and Reuters coverage, says the draft terms would have shifted much of the legal risk for AI errors onto business customers. Zetik is an aggregator, and USA 24 could not confirm that claim in a primary document, so it should be treated as reported rather than established.

Several questions remain open. Meta has not announced a new implementation date or confirmed whether it will rewrite the disputed provisions. It is also unknown whether the Norwegian injunction and the CMA probe will create binding limits on Meta outside those jurisdictions. Whether any liability-shift language would have applied to US advertisers has not been confirmed.

Liability language is a live issue across AI products. USA 24 recently reported on xAI's Grok bot, which now calls US businesses and leaves users liable for what happens on those calls.

On the US side, the federal picture is thin. The reporting reviewed here found no US federal or state action tied to Meta's proposed terms. Broader questions about which rules bind AI labs are being debated separately, as USA 24 has covered in its look at the Super Intelligence Task Force's plan to list federal laws binding AI labs.

What US publishers and businesses should do now

• Check which websites are linked to your Facebook or Instagram business accounts, and remove any connections you no longer need.
• Read any Meta terms update notice closely before accepting. Note the effective date, which document is changing, and whether the update adds AI training or content-use language.
• If you run a publication or blog, document suspected scraping with server logs and dated notes. Records make it far easier to raise a complaint.
• Contact your trade association. The Norwegian case was backed by an industry association and 180 companies, which shows collective action can carry weight.
• Advertisers and small businesses should review any contract notice before accepting it, especially if it adds liability terms.

These are general precautions, not legal advice. The reporting reviewed here found no US regulatory filing on this dispute.

What to watch next

Three developments will matter in the coming weeks. The first is whether Meta publishes a revised version of the terms, and whether that version addresses the linked-domain language that drew the sharpest criticism. The second is how the Norwegian court proceeds beyond the temporary order, and what the CMA does after its announced review.

The third is whether a US regulator or state attorney general acts, or whether lawmakers follow Schiff's lead. The next move belongs mostly to Meta, which must decide whether to rewrite the language, reintroduce a narrower version or let the proposal fade. Publishers, advertisers and small businesses will be watching for the first revised notice, and they should read it closely when it arrives.