October 18, 2026, (Inside AI) — Meta has removed a controversial clause from its commercial terms of use after legal experts warned it could allow the company to scrape content from websites, including news sites, to train its artificial intelligence models. The provision, which appeared on Meta's commercial terms page, stated that Meta may use any content from domains, web pages, or URLs to "provide, develop, improve, research, train, and maintain the integrity of" its products and services. The clause vanished from the page on Saturday, October 17, following widespread criticism.
The incident highlights growing tensions between AI companies and content creators over data used for training. Legal experts, including James Grimmelmann, a professor of digital and information law at Cornell University, argued that the now-deleted provision would have granted Meta the right to use scraped content for free and override blocking instructions in a site's robots.txt file. Robots.txt is a standard file that websites use to tell web crawlers which pages they can or cannot access.
Meta's spokesperson defended the update, saying it was a consolidation of existing terms and did not change how the company crawls the web. "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 said. Andy Stone, Meta's vice president of communications, added: "There is no change to how Meta uses web crawlers; any suggestion to the contrary is false."
The swift reversal underscores the scrutiny tech giants face as they expand AI training data collection. Meta already scrapes publicly shared content from users above 18 for AI training, a practice that has drawn privacy concerns. Similarly, Elon Musk's xAI uses public posts on X to train its Grok chatbot. Both companies do so without obtaining user consent first, a point of contention among regulators and users.
Read: ChatGPT product liability
This is not the first time a tech company has quietly updated its terms to allow AI training. In July 2025, file-sharing service WeTransfer revised its terms to remove mentions of machine learning and AI after backlash. The original update suggested user files could be used to "improve machine learning models." WeTransfer had previously only claimed rights to "reproduce, modify, distribute and publicly display" content.
Legal Battles Over AI Training Data Intensify
Meta's removal of the clause comes amid ongoing legal battles over AI training data. Last year, Meta defeated a high-profile copyright infringement lawsuit, but similar cases against AI companies continue. The outcome of these cases could set precedents for how AI firms collect and use data. Legal experts argue that terms of service should not override robots.txt, which is widely respected as a signal of content owners' preferences. The now-deleted Meta provision would have effectively bypassed that norm, raising questions about the balance between AI innovation and content rights.
Meta's quick retraction may reflect an awareness of the reputational and legal risks. The company has faced criticism for its data practices, and the backlash over this clause shows that users and experts are watching closely. As AI models require ever-larger datasets, companies are under pressure to find ethical and legal ways to source data. Some are turning to licensed content or synthetic data, but scraping remains common.
The incident also highlights the role of public pressure. Within hours of the clause being flagged by experts, Meta removed it. This suggests that transparency and public scrutiny can influence corporate behavior, even in the fast-moving AI industry.
Meta's terms of use are not the only ones under scrutiny. Other tech companies have faced similar backlash. For instance, Google and OpenAI have been criticized for scraping content without clear consent. The European Union's AI Act and other regulations are beginning to address these issues, but enforcement remains uneven.
As AI continues to evolve, the debate over data rights will likely intensify. Meta's reversal may be a small victory for content owners, but the larger battle over AI training data is far from over. The company's clarification that robots.txt preferences will be respected is a reassurance, but experts will be watching to see if Meta's actions match its words.
Read: Anthropic AI data breaches
In the meantime, content owners are advised to monitor their robots.txt files and terms of service. The incident serves as a reminder that in the age of AI, vigilance is necessary to protect intellectual property.
