If you manage a website and have a Meta business account, this affects you directly. The company has modified its business terms following criticism from the publishing sector, but several legal experts maintain that the scope for continuing to use web content in the development of its artificial intelligence remains open.
The new terms, which go into effect on October 30, no longer include the word 'train.' However, the rest of the language related to content usage remains practically intact, according to an analysis published by The Verge and also covered by Nieman Lab.
What the original text said and what Meta removed
The initial version of the terms for business accounts was explicit: Meta could use any content from websites associated with a business account to develop, research, and train its products.
For a digital media outlet with an advertising presence on Meta's platforms, that wording opened a door that many publishers had not considered when accepting the terms of their business account.
Following the reaction from the publishing sector, Meta removed the direct reference to training with crawled content. The rest of the clause, however, remains in effect, and it is precisely that remainder that worries the specialists consulted.
The objection of legal experts: 'develop' can also mean 'train'
James Grimmelmann, a professor of digital law at Cornell Law School, was clear when assessing the change to Nieman Lab. As he explained, the modification does not restrict the expression 'associated with its products or services,' which would allow Meta to crawl entire websites.
Matt Topic, a lawyer specializing in media and intellectual property at the firm Loevy & Loevy, points out another weak point. The revised version continues to allow the use of content to 'develop any business product,' a term that is not defined in the text.
Topic maintains that Meta develops business products based on artificial intelligence, so, in his opinion, the word 'develop' is broad enough to include what was previously called 'train.'
The tension between business terms and the robots.txt file
The underlying conflict combines technical and legal aspects. The robots.txt file is the tool that websites use to indicate to crawlers, including those from search engines like Google, which content they can crawl and which they cannot. This is a historical convention, not a binding contract.
The experts consulted by Nieman Lab warned that the business terms could allow Meta to argue that its conditions prevail over other blocking instructions. In practice, this means that a publisher who blocked crawling via robots.txt could find that Meta claims to have obtained permission through another route: the acceptance of business account terms.
Meta’s position: it is not about crawling news sites
Meta rejects that interpretation. Its spokesperson, Daniel Roberts, told The Verge that content owners can use robots.txt to signal their preferences. He added that the update is not intended to crawl news sites and that the changes do not grant new rights that override those preferences.
This is a statement of intent by the company, not a contractual clause. The legal experts cited do not question the intention expressed by Meta, but rather what the contract text allows it to do. Between what the company says it will do and what the contract authorizes it to do, there is a margin that, for the moment, has not been closed in writing.
What changes for publishers and advertisers with a business account
The practical effect falls on those who have a business account on Meta and, at the same time, manage their own website. A digital media outlet that manages its page and advertising campaigns on the company's platforms could be granting, by accepting the terms, permissions regarding its content that it never intended to cede.
There is a relevant nuance: the wording refers to sites 'associated with an account,' and it is precisely that association that, according to Grimmelmann, is not sufficiently defined in the text. Small media outlets, without their own legal department, are the ones with the fewest resources to detect the real scope of this clause.
What to check before October 30
Based on what the consulted legal experts pointed out, these are the points that should be checked before the new terms go into effect.
Review the full text of the new terms and locate the expression 'associated with its products or services,' considered the most problematic by Grimmelmann.
Check which websites are linked to the Meta business account. The broader that association, the greater the potential scope of the clause.
Keep the robots.txt file updated. Meta claims it will respect those preferences, and documenting them reinforces the publisher's position in the event of a dispute.
Consult with a legal advisor if the site contains paid content or original editorial material, as the term 'develop' could be interpreted more broadly than 'train.'
Three possible scenarios from now on
October 30 marks the date from which continuing to use Meta's business tools may be equivalent to accepting the terms as written, with the ambiguity pointed out by Grimmelmann and Topic.
From here, three paths open up. That Meta clarifies in writing the scope of expressions such as 'business product' and 'associated.' That it maintains the current text and relies on its public statement regarding robots.txt. Or that a publisher decides to take the interpretation of these terms to court.
Until one of these scenarios occurs, the guarantee available to publishers is a public promise from the company, not a closed contractual clause.
Frequently asked questions
What exactly did Meta change in its business terms?
Meta removed the word 'train' from the clause describing the use it can make of content from websites associated with a business account, but kept expressions such as 'develop' and 'associated with its products or services.'
When do Meta’s new terms go into effect?
The new terms for business accounts go into effect on October 30.
Why do legal experts believe Meta could still use content for AI?
Because they consider that the term 'develop,' which remains in the text, is broad enough to encompass what was previously called 'train,' as James Grimmelmann and Matt Topic explained to Nieman Lab.
Does the robots.txt file protect publishers against Meta?
Meta claims it will respect the preferences indicated in robots.txt, but the business terms could allow it to argue that its conditions prevail over those blocking instructions, according to the experts consulted.
Who is affected by this change in terms?
It mainly affects those who manage their own website and, at the same time, have a business account on Meta's platforms, especially digital media and advertisers with an advertising presence on the company.
What does Meta say about this criticism?
Its spokesperson, Daniel Roberts, told The Verge that the update is not about crawling news sites and that the changes do not grant new rights that override the preferences marked in robots.txt.
What should publishers do before October 30?
Review the full text of the terms, check which websites they have linked to their business account, update their robots.txt file, and consult with a legal advisor if their content is paid or original editorial material.
