Meta Revises Commercial Terms of Service Amid AI Scraping Concerns

21 October 2026 3 min read هدهد تك | Hudhud Tech

Meta has updated its commercial terms of service, removing the word “train” from its policies following criticism regarding its AI scraping practices. The revised terms will take effect on October 30.

Last week, legal experts raised concerns that Meta’s original terms appeared to allow the company to scrape content from websites linked on business pages without restrictions. This change was seen as a potential way for Meta to avoid paying licensing fees to news organizations by requiring business users to provide data for AI training simply by maintaining a business page.

A Meta spokesperson confirmed that the company did not inform users about the language changes, which could lead to further confusion as the new policy approaches. In response to the backlash, the spokesperson stated, “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.”

However, experts remain skeptical about the updates. Matt Topic, a media and intellectual property attorney, expressed that the revised terms still grant Meta broad permissions to use scraped content for developing business products, which could include AI applications. He noted that the term “develop” is sufficiently vague to encompass “train.”

The initial version of Meta’s terms allowed the company to use “any content” from business pages for various purposes, including training and maintaining its products. The updated terms now permit Meta to “periodically pull updates” from business users’ websites to support advertising and business products, but they no longer mention the use of “any content” or the word “train.” Meta claims this scraping is intended to benefit users by enhancing its research into products that help create and deliver ads.

Despite the changes, experts like James Grimmelmann from Cornell Law School argue that the language remains overly broad and could still allow extensive scraping. He pointed out that the terms still permit Meta to scrape content associated with users’ products or services.

Legal experts suggest that a more effective solution would be for Meta to include explicit language in its policies to respect robots.txt files, which indicate a website’s preferences regarding automated scraping. Since compliance with robots.txt is voluntary, the absence of enforceable language could allow Meta to ignore such preferences.

Criticism of Meta’s terms extends beyond legal circles, with news organizations worldwide urging regulators to intervene. Theo Bamber, CEO of the UK News Media Association, condemned Meta’s actions as a “shameless land grab” and called for regulatory bodies to protect creators and businesses from what he described as bullying tactics.

Similarly, the European Publishers Council has appealed to the European Commission for intervention, emphasizing that publishers should have control over how their journalism is utilized and compensated. Danielle Coffey, president of the News/Media Alliance, labeled the terms as “outrageous and probably illegal,” urging U.S. regulators to recognize the exploitative nature of Meta’s policies.

Coffey concluded that Meta’s strategy to withhold access to its audience unless businesses allow scraping is a clear abuse of market power that must be addressed before the policy takes effect on October 30.