Meta removes language that may have allowed it to scrape news sites

Meta has quietly updated its commercial terms to strip out language that appeared to grant it broad rights to scrape content from third-party websites, including news outlets, for the purpose of training its artificial intelligence systems. The change comes after criticism from publishers and other stakeholders who objected to the implied permission to harvest their content without explicit consent.

The removed clause had stated that Meta could use content from sites that opted into its commercial terms for AI training purposes. By taking the language down, Meta is narrowing the scope of what it can claim as permissible data collection under its existing agreements, though it does not necessarily eliminate all scraping capabilities.

This development highlights the growing tension between large AI companies and content creators over data rights. As AI training datasets become increasingly dependent on web-scraped content, publishers and rights holders are pushing back against terms that appear to grant sweeping access without compensation or opt-in mechanisms.

The move may set a precedent for how other tech companies structure their data collection terms going forward. It also underscores the regulatory and reputational pressure that AI firms face as governments and the public scrutinize the sourcing of training data.

This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.