Meta retains broad access to business websites
The company removed references to AI training from its new terms, but largely retains the authority to crawl websites.
Listen to this article
There is no audio version yet. Request one and an AI voice will read the article aloud.
Read by an AI voice.
Meta has amended the wording of its commercial terms following criticism of a provision on the automated collection of website content. The reference to training products has disappeared, but permission to crawl business users’ websites remains broadly worded.
The new terms are due to take effect on 30 October. They apply to businesses using Meta services, for example for advertising, Facebook pages or other business functions. The original version stated that Meta could collect and use content from linked websites for purposes including developing and training its products.
Following criticism from publishers and legal experts, Meta removed the reference to training. The amended text links crawling to advertising and business products, product development, research, and the safety and integrity of Meta’s services.
The change does not resolve every question. The terms still give Meta permission to use automated software to index, analyse and periodically revisit domains, web pages and URLs. The publication also says that the provision allowing the company to create test data to gain access to websites remains in place.
Meta says the terms do not introduce a new practice and relate only to advertising and business products. The company also says that website operators can use robots.txt to indicate which crawlers should not be granted access.
Legal experts consulted by Nieman Lab consider the wording broader than Meta’s explanation. They point out that words such as ‘develop’ and ‘improve’ do not specify precisely which forms of reuse are permitted. This does not automatically mean that Meta thereby obtains the right to use journalistic content for AI training; the legal uncertainty described does not settle that question.
For Dutch businesses, the permission described raises practical questions: which linked domains are covered, which crawler rules apply there and what do those settings mean for Meta’s access?
The change is therefore not a complete withdrawal of the scraping policy, but an adjustment to the stated purposes. What Meta may do with specific types of website content will depend on the final terms, the technical implementation and any proceedings before regulators or courts.
Also show the region and a close-up
One story, several perspectives
What is established
- Meta is amending its commercial terms from 30 October.
- The explicit reference to training has been removed from the amended text.
- Automated crawling of websites linked to business services remains permitted under the terms described.
- Meta says that robots.txt can indicate preferences.
Left
Arguments Platforms should not gain access through general terms and conditions to journalistic or business content that is subsequently used commercially. Publishers and creators deserve control and a fair licensing model.
Values Copyright, the balance of power and protection of public information.
Consequences Stricter rules could slow innovation, but would prevent large platforms from collecting content without clear permission or compensation.
Centre
Arguments Crawling may legitimately be necessary for advertising, product information and security, but the terms should clearly define which content is used for what purpose. Transparency and oversight matter more than broad claims.
Values Proportionality, legal certainty and workable digital markets.
Consequences Clear technical and legal boundaries could protect both platform functions and the interests of website operators.
Right
Arguments Businesses choose voluntarily to use Meta’s business services and, according to Meta, can restrict access through robots.txt. Excessively heavy regulation could deprive small businesses of the marketing channels on which they rely.
Values Freedom of contract, entrepreneurship and limited government intervention.
Consequences Broad permission may deliver greater efficiency and better advertising, but businesses would then bear more responsibility for their access settings.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved with corrections · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The amended terms and the continuation of crawl rights have been described independently. The text avoids the unproven conclusion that Meta will now certainly use the content for AI training.
- confirmed Meta has removed the explicit reference to training from the amended crawling clause. — Reported by Search Engine Journal and TechWyse. source
- confirmed The new terms are due to take effect on 30 October. — The publications refer to the announced effective date. source
- incorrect The remaining wording proves that Meta will use website content for AI training. — The sources describe legal uncertainty, not proven use. source
1 correction(s) applied
- Was: Meta may scrape users’ websites for AI training.Now: Meta retains broad crawl rights; the explicit reference to AI training has disappeared from the amended terms, but the legal scope remains a matter of debate. (The original wording presented an uncertain legal consequence as an established fact.)
Editor's note
The removal of the explicit reference to training and the continuation of broad crawl rights have been reported by several trade publications. The legal consequences and the precise scope of ‘develop’ and ‘improve’ remain uncertain.Sources
- Meta Drops Training From Its Crawl Clause And Keeps The Crawl — Search Engine Journal
- Meta Narrows Its Website Crawling Terms — TechWyse
- Meta’s new commercial terms will allow the scraping of news sites, legal experts say — Nieman Journalism Lab
More on this in Dutch media
- NU.nl — „meta facebook”
- De Telegraaf — „meta facebook”
- de Volkskrant — „meta facebook”