Meta Drops 'Train' From Commercial TOS After AI Scraping Backlash
Meta has adjusted its upcoming commercial terms of service to remove the word 'train', though legal experts warn that sweeping scraping permissions remain in place.
Radwan AbdulhadiEditor and technology writer
Meta Removes 'Train' From Updated Commercial Terms
In an effort to quell mounting criticism, Meta has quietly dropped the word “train” from its "streamlined" commercial terms of service, which are scheduled to take effect on October 30. Reported by Ashley Belanger, the policy update follows significant backlash from legal experts who initially warned that the tech giant's sweeping new commercial terms could permit Meta to endlessly scrape websites linked on business pages.
Legal experts previously warned that Meta could potentially leverage the policy to bypass millions of dollars in licensing fees owed to news organizations. By requiring all business users to share vast amounts of AI training data simply for operating a business page, the initial language sparked immediate concern across the industry.
Meta Confirms Changes and Defends Policy Scope
A Meta spokesperson confirmed that the company did not directly notify users about the modified language following the initial backlash, a move that critics suggest may contribute to further confusion ahead of the rollout. In a statement provided to Ars Technica, Meta acknowledged earlier reports but maintained that the policy updates do not expand its scraping rights.
“We’ve heard you and want to clarify: the crawling terms we consolidated only applied to our ads and business products, and we’ve clarified that in the terms themselves. It does not change how Meta uses web crawlers. As always, content owners can use robots.txt to signal their preferences. This is not about scraping news sites, and these updates don’t give us new rights that supersede those preferences,” the spokesperson stated.
Legal Experts Remain Skeptical of Revised Policy
Despite the removal of the word "train," the same legal experts who sounded the initial alarms told Nieman Lab that the updates do not go far enough to shield business page users from invasive web scraping. Matt Topic, a media and intellectual property attorney at Loevy & Loevy, noted that publishers should continue to express concern over the wording.
Initially, Meta’s terms permitted the company to utilize any content from websites posted on business pages to provide, develop, improve, research, train, and maintain the integrity of any Meta products or features. The updated terms specifically delete the word “train” and remove references to "any content."
Instead, the revised guidelines allow Meta to periodically pull updates from business users’ websites to develop and improve advertising and business products, as well as to maintain the security of its infrastructure. However, Matt Topic pointed out that the term "develop" remains broad enough to implicitly include training activities for new AI products.
Debate Over Robots.txt and Enforceable Preferences
James Grimmelmann, a professor of digital and information law at Cornell Law School, acknowledged that Meta added clearer language regarding permitted uses. However, he told Nieman Lab that the phrase allowing Meta to develop advertising and business products is still broad enough to drive a truck through, failing to adequately restrict comprehensive site scraping.
University of Wisconsin law professor Benjamin Sobel suggested that a truly effective solution would require Meta to explicitly state in the policy that the company agrees to respect scraping directives outlined in robots.txt files. Because compliance with robots.txt is ordinarily voluntary, experts argue that legally enforceable terms of service omitting mentions of robots.txt could still allow Meta to bypass blocks implemented by site owners.
Global News Groups Call for Regulatory Intervention
Criticism of the commercial terms extends far beyond initial reporting, with global news groups actively urging regulators to intervene and block the implementation of Meta's policies. Theo Bamber, chief executive for the United Kingdom News Media Association, labeled the move a shameless land grab driven by platform bullying, calling on UK regulators to protect creators.
Similarly, the European Publishers Council has officially asked the European Commission to examine the situation. Executive Director Angela Mills Wad emphasized that publishers must retain the right to determine how their journalism is utilized and negotiate fair compensation.
In the United States, Danielle Coffey, president and CEO of the News/Media Alliance, condemned the terms as outrageous and likely illegal. Coffey urged US regulators to recognize the policy as an abusive exercise of market power that should be halted ahead of its October 30 launch.
Sources
- Ars TechnicaMeta drops the word "train" from its commercial TOS after AI backlash
Continue chronologically
Related entity coverage
Related stories
Meta Delays Controversial New Business Terms Amid Scraping Backlash
Meta has pushed back the effective date of its controversial new commercial terms to November 20 following widespread scraping concerns and industry pushback.
Meta Updates Terms, but AI Scraping Concerns Remain
Legal experts warn that Meta's revised business terms still leave room for the company to scrape and train AI models on news publisher websites.
Meta Removes Language That Raised Scraping and AI Training Concerns
Meta has removed controversial language from its commercial terms of use following concerns regarding website scraping and AI model training.
Meta Removes Language That May Have Allowed AI News Scraping
Following public backlash and legal expert interpretation, Meta is updating its terms again to remove controversial language about web crawling and AI training.
