Revised terms from Meta would give the platform authority to gather content from third-party websites connected to business accounts, drawing that material into the development and refinement of its AI systems — a provision that has drawn pushback from outside the company. Slated to come into force on 30 October 2026, the adjustments have drawn questions from publishers and legal professionals around issues of copyright, consent, and who ultimately holds authority over digital content.

Organisations that rely on Meta's commercial tools — among them Facebook Pages, Instagram business accounts, and Meta Business Suite — fall within the scope of the revised terms. Under the updated agreement, Meta is permitted to deploy automated systems — web crawlers, bots, and similar tools — to gather, examine, and periodically pull content from websites tied to businesses and the products or services they offer. Under those terms, data gathered by the platform can be fed into the development, refinement, and training of Meta's products.

Meta has described the revisions as a move toward clearer, more streamlined rules, yet attorneys who reviewed the language noted that the phrasing may open a pathway for the company to access website content on a scale publishers had not anticipated. Given that Meta operates a range of AI-driven products — among them its Meta AI assistant — the wording of its terms has drawn scrutiny over whether published articles, photographs, and other online content might feed into the company's model-training processes.

News publishers find themselves among the most directly affected parties, given that their platforms depend heavily on original journalism and content protected under copyright law. Licensing deals between news publishers and technology firms — covering articles and other produced content — represent a routine part of how the industry operates. Some observers have raised concerns that Meta's updated terms of service may open a pathway for gathering comparable content without entering into separate licensing deals with individual rights holders.

Some attorneys who follow the field have raised doubts about whether the updated terms might be read as superseding site-level directives that bar automated web crawlers from accessing content. Outlined within the document is a provision for generating test credentials, a detail that has prompted questions about how platforms gate content behind user registration. Whether those provisions would withstand judicial scrutiny is an open question.

Litigation now under way between publishers and technology companies centres on a dispute that has yet to be resolved: whether feeding copyrighted material into AI training systems falls within the bounds of the law. The company's updated terms, it has said, are not intended to serve as a legal foundation for overriding robots.txt directives that block automated data collection. Documentation is also available from the firm for site operators seeking to prevent its web crawlers from accessing their pages.

Those who reviewed the agreement's language noted that its imprecise phrasing may leave certain provisions open to interpretation. Businesses that do not agree to Meta's updated commercial terms stand to lose all access to the company's products for commercial purposes — a consequence that goes beyond simply forfeiting the extra permissions the new agreement would have unlocked.

At the heart of the disagreement lies a widening tension between the pace of AI advancement and the authority publishers hold over their own content. AI developers depend on access to large collections of data to build and improve their systems, while publishers are seeking greater control over how their content is reproduced and circulated by others. Any dispute over how Meta reads those contractual terms carries the potential to reshape the rules governing online businesses and digital media more broadly.

Questions about how AI companies gather and repurpose written material have left publishers without a clear path forward.