Meta is revising its commercial terms to remove language that was interpreted as granting it the right to scrape content from external websites, including news sites, to train its AI models. The change follows public backlash and legal commentary suggesting the wording could have been used to justify copying content without permission.
Background
AI companies have faced growing scrutiny over their use of publicly available web content to train large language models, with news publishers and authors increasingly pushing back over copyright and compensation concerns.
- Source
- The Verge
- Published
- Oct 17, 2026 at 11:04 PM
- Score
- 5.0 / 10