What Meta’s New Terms Meant for News Website Scraping

What did Meta’s terms appear to allow?

Meta’s commercial terms cover business accounts used to manage services such as Facebook pages, Instagram profiles, Messenger accounts, and WhatsApp channels. Under the version described by Nieman Lab, Meta could use automated tools to visit websites and gather content from sites connected with those accounts. ¹

The terms were due to take effect on October 30. Nieman Lab reported that businesses continuing to use Facebook, Instagram, or WhatsApp would be treated as accepting them. News organizations were among the businesses that could be affected because they commonly have business accounts on Meta’s platforms. ¹

This was a description of what the terms seemed to permit, not evidence that Meta had already collected material from a particular publisher. The reporting does not identify specific sites or pages that were accessed. ¹²

What kind of website content was at issue?

The reports discuss material on websites associated with business accounts, but do not provide a detailed inventory of what Meta might gather. They do not say that every page on a publisher’s site would necessarily be collected, or explain how Meta would decide which pages were connected to an account. ¹²

The wording removed from the terms referred to using material from relevant websites in connection with Meta’s products, services, and features. It listed several possible purposes, among them research, product improvement, and training. The Verge reported that experts understood this language as potentially covering AI training. ²

The sources do not explain whether the wording would have applied only to publicly viewable pages, how often any collection might occur, or what safeguards would have applied. Those details cannot be inferred from the reported language alone. ¹²

Did Meta actually scrape news sites?

The reporting describes what the terms appeared to permit, not proof that Meta had collected content from particular news sites. The sources provided do not identify any gathered articles, describe a specific scraping operation, or establish that news content was used to train an AI system. ¹²

That distinction matters. A contract term may describe a claimed right or permission, but it does not by itself show that a company exercised it. The reports also do not settle the legal meaning of the wording or how it would apply to individual publishers. ¹²

Matt Topic, a media and intellectual-property attorney quoted by Nieman Lab, questioned the scale of access the provision could create and compared the material to content that might otherwise be licensed. His comment reflects a concern about the terms; it is not evidence that such collection or use occurred. ¹

What changed after the concerns were reported?

On October 17, The Verge reported that Meta had removed the disputed language from its commercial terms. The removed passage had referred to using website material for a range of purposes, including training and research. ²

The reports do not supply the complete replacement text or explain whether other parts of Meta’s terms still address website content. So the removal is a meaningful change to the wording identified by The Verge, but the available information does not answer every question about Meta’s policies or practices. ²

The timing is notable: Nieman Lab reported that the commercial terms were set to take effect on October 30, while The Verge reported the removal on October 17. The sources do not say whether the revised terms were final or whether Meta would make further changes. ¹²

What could this mean for publishers and readers?

For publishers, the concern was that material on their websites could be collected and used under terms connected to their business accounts, potentially without a separate licensing arrangement. Nieman Lab quoted Topic saying that the provision might give Meta access to material that could be expensive to license. ¹

The reports do not establish whether any publisher’s content was collected, whether publishers received notice, or whether they had a way to opt out of the disputed use. They also do not explain what recourse a publisher might have. Those questions remain open in the sources provided. ¹²

For readers, the reporting does not establish a direct change to what appears on Meta’s platforms or how individual users’ data is handled. The issue described concerns material on websites linked to business accounts. Whether any news content was gathered or used, and what the revised terms mean in practice, remains unclear. ¹²