Key points
- Meta paused the changes to engage with stakeholders and review the terms.
- Preliminary injunctions from courts in Brazil and Norway halted the update.
- The UK Competition and Markets Authority opened a formal investigation into the move.
Meta has indefinitely delayed its planned update to commercial terms, which would have mandated that Facebook page owners allow extensive AI training on external websites linked from their pages.
The claim
The updates, originally set for October 30 and delayed to November 20, would have required businesses to permit Meta to crawl linked external websites for AI training.reported According to Ars Technica, the terms offered no opt-out mechanism for businesses, including news publishers involved in legal battles over unwanted AI scraping.reported The revision also mandated that publishers provide test login details to enable Meta’s automated systems to consistently circumvent paywalls. Lawyers cautioned Nieman Lab that ambiguous wording might permit Meta to disregard robots.txt scraping directives and extensively scrape content outside of Facebook.
Evidence: reported
“Does Meta think we can’t read?” Meta abandons sketchy commercial terms update.aifleet-ars
Evidence: reported
“Does Meta think we can’t read?” Meta abandons sketchy commercial terms update.aifleet-ars
The response
A Meta representative stated to Reuters that the company is halting the updates to consult with stakeholders and ensure the changes are implemented correctly, while current terms stay in place. UK News Media Association chief Theo Bamber stated on LinkedIn that replacing “train” with “develop” in Meta’s terms failed to comfort publishers because the new term carried the same risk. He argued that the company could use all scraped content to develop any business product.
The bottom line for teams
The update faced pushback from users, courts, and regulators globally. Legal bodies in Brazil and Norway issued temporary orders to halt the terms following lawsuits brought by organizations. The UK’s Competition and Markets Authority launched a formal probe to determine if Meta was leveraging its dominant market position to acquire valuable AI training data. Nieman Lab highlighted bipartisan worry in the United States that Meta could sidestep millions of dollars in licensing costs.
What the paper shows
The report examines whether Meta’s updated commercial terms would legally compel businesses to surrender control over external content linked from their Facebook pages for AI training. The findings indicate that the terms lacked opt-out mechanisms and required credential sharing for paywall bypassing. The analysis highlights vague language that could override robots.txt preferences, raising concerns about exhaustive crawling of off-Facebook content without explicit consent or licensing agreements.
Why it matters
- Model evaluators must assess whether Meta’s scraping practices comply with existing robots.txt standards and licensing norms.
- The lack of opt-out mechanisms in the proposed terms highlights potential risks for businesses relying on external content for their pages.
- Regulatory investigations in the UK and injunctions in Brazil and Norway signal growing scrutiny of AI training data acquisition methods.
- The requirement for test credentials to bypass paywalls raises ethical and legal questions about content access and intellectual property rights.
Topics: AI training, Meta, commercial terms, regulatory pushback, data scraping
Updated
Then / now / next
Meta planned to enforce new commercial terms requiring businesses to allow AI training on linked external content.
Meta has indefinitely paused the update following legal and regulatory pushback.
Teams should monitor ongoing investigations and potential further legal actions against Meta’s AI data practices.
Story timeline
- Then Meta planned to enforce new commercial terms requiring businesses to allow AI training on linked external content.
- 06 Nov 2026 “Does Meta think we can’t read?” Meta abandons sketchy commercial terms update.
- Now Meta has indefinitely paused the update following legal and regulatory pushback.
- Next Teams should monitor ongoing investigations and potential further legal actions against Meta’s AI data practices.
The strongest doubt
Counterpoint from the sources
Meta argues that pausing the changes allows time to engage with stakeholders and ensure the terms are correct, suggesting a willingness to address concerns rather than dismiss them outright.
For readers
Policy
- Regulators in multiple countries are scrutinizing Meta’s AI data acquisition practices.
- Courts in Brazil and Norway have issued injunctions against the proposed terms.
Business
- Businesses were required to allow AI training on linked external content with no opt-out.
- Publishers were asked to provide test credentials to bypass paywalls for Meta’s bots.
Researchers
- The vague language in the terms could override robots.txt preferences.
- Nieman Lab reported potential for Meta to avoid millions in licensing fees.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Questions readers ask
What were Meta’s proposed changes to commercial terms?
Meta intended to mandate that Facebook page owners permit AI training on linked external sites, offering no way to opt out.
Why did Meta pause the update?
Meta paused the changes to engage with stakeholders and address backlash from users, courts, and regulators.
What actions have regulators taken?
Courts in Brazil and Norway issued injunctions, and the UK’s Competition and Markets Authority opened an investigation.
Connected stories
Reading receipt
- Meta paused the changes to engage with stakeholders and review the terms.
- Preliminary injunctions from courts in Brazil and Norway halted the update.
- The UK Competition and Markets Authority opened a formal investigation into the move.
Still open: Will Meta revise the terms to include opt-out mechanisms for businesses?
Still open: How will ongoing investigations in the UK and other regions impact Meta’s AI data practices?