Meta removes language that may have allowed it to scrape news sites
Meta updates its commercial terms to remove ambiguous language after critics argued it permitted scraping news content to train AI models.
Stock photo for illustration only, not from the actual event
- Meta removes controversial commercial terms criticized for allowing news scraping.
- Experts warned the language permitted free data collection to train AI models.
- Communications executives defended web crawler policies before the update.
- The company clarified that site owners retain full control via robots.txt.
Meta has updated its terms of service once again following widespread backlash over phrasing that suggested the tech giant could harvest content from news websites and other online domains to train its artificial intelligence systems. The disputed text, which previously appeared on the company's commercial terms page, has now been completely removed.
Legal scholars and media analysts, including Neiman Lab and James Grimmelmann, a digital and information law professor at Cornell University, had previously interpreted the phrasing as an attempt by Meta to grant itself legal rights to scrape site content for AI training without compensation, potentially overriding the blocking preferences outlined in a website's robots.txt file.
Stock photo for illustration only, not from the actual event
Stock photo for illustration only, not from the actual event
In response to the initial concerns, Meta communications executives Andy Stone and Daniel Roberts pushed back against the criticism, insisting that there were no alterations to how the company's web crawlers operate and dismissing contrary suggestions as false. However, the latest update goes a step further by eliminating the controversial passage altogether.
“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.”
Daniel Roberts
In a follow-up statement provided to industry media, Roberts emphasized that the revised terms explicitly separate ad and business products from general crawling behavior. He reiterated that content owners maintain the ability to signal their preferences using robots.txt files and that the policy changes do not grant Meta any new rights that supersede those instructions.
This policy adjustment highlights the intense scrutiny technology companies face regarding data acquisition practices for generative AI development. As creators and publishers demand stricter protections for their intellectual property, legal frameworks and platform terms are undergoing rapid re-evaluations to address transparency and consent in data harvesting.
Source: The Verge
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment