Meta drops the word "train" from its commercial TOS after AI backlash
Meta has modified its upcoming commercial terms of service to remove the word “train” following heavy criticism from legal experts and publishers.
Meta has modified its upcoming commercial terms of service to remove the word “train” following heavy criticism from legal experts and publishers. The policy update, which is set to take effect on October 30, originally sparked warnings that the tech giant intended to scrape business website content to avoid paying millions in artificial intelligence licensing fees.
A Meta spokesperson confirmed the quiet policy change to Ars Technica, clarifying that the rules apply exclusively to ads and business products and do not alter how web crawlers function. Despite the modification, legal scholars argue the updated text remains overly permissive. Intellectual property attorney Matt Topic noted that terms permitting Meta to "develop" advertising and business products still effectively allow for training.
Legal experts say the revised terms fail to incorporate explicit protections honoring robots.txt files, which website owners use to block automated scraping. Without legally enforceable language acknowledging these scraping preferences, critics warn that Meta can still bypass creator intent.
Global news organizations are continuing to push back against the upcoming rollout. Trade associations including the European Publishers Council, the UK News Media Association, and the US News/Media Alliance have urged regulators to intervene and halt the October 30 implementation.
A concise explanation prepared from the supplied report material. It is a sourced brief, not a claim of original reporting.
The publisher link, attribution and available publication details are preserved so you can check the original report.
3 related WireUnWired pages are available below for additional context.
Original publication: 21 October 2026 01:37
