EDITION OF OCTOBER 23, 2026
Meta delays commercial terms after publishers raise AI objections
Meta is postponing the implementation of its new commercial terms until November 20 following publishers’ objections over their possible implications for AI training. The announced delay is intended to allow further explanations, but does not resolve questions about the use of content.
Legal disclosure: this content was produced by an artificial intelligence system. Editorial transparency
Meta’s announced postponement leaves open the central question raised by publishers: what uses of their content would the new commercial terms permit? The promised explanations about web crawlers and the robots.txt file will need to be assessed in light of this question about the scope of the terms.
Meta is postponing the implementation of its new commercial terms to November 20 from the originally planned October 30. In a statement to The Verge, spokesperson Daniel Roberts explains that the delay is intended to allow the company to better explain the changes to affected partners. The dispute concerns the possibility that these terms could authorize the collection of content from news websites to train artificial intelligence systems.
According to The Verge, the original wording appeared to give Meta this permission for websites using its platforms. The company subsequently removed a reference to training its products. However, legal experts cited by the outlet believe this change is insufficient to rule out that possibility. The disagreement therefore concerns the scope of the terms, beyond the presence or absence of an explicit term.
Opposition has taken shape among news industry organizations. The News/Media Alliance sent Mark Zuckerberg a letter on Thursday requesting a revision of the text. Publishers and digital advertising industry participants in the European Union and the United Kingdom have also asked regulators to require the withdrawal of these new terms. The postponement comes amid this collective pressure.
Meta says it intends to use the delay to explain how its web crawlers work. The company also says that content owners will still be able to express their preferences through the robots.txt file. This response describes a mechanism for signaling preferences, but does not, on its own, establish how each use of content will be governed by the disputed terms.
From an analytical perspective, the case highlights two related questions: technical access to pages and the permissions associated with their use. Clarifying how a crawler works can help publishers understand collection. That does not necessarily resolve a contractual ambiguity over the uses permitted after collection. The removal of the training reference and the continued legal objections illustrate this difficulty.
At this stage, the established change is a delay in implementing the terms. The article reports neither an agreement with the organizations opposing the terms nor revised wording that resolves their objections. It also provides no basis for concluding that Meta has actually used publishers’ content to train its models on this basis.