Legal experts say Meta's revised terms still allow AI training on scraped news
Meta removed the word "train" from its business terms after publisher backlash, but lawyers tell Nieman Lab the update still permits scraping and AI training.
Meta's revised business terms, which take effect October 30th, still allow the company to scrape news sites and train AI on that content, according to legal experts cited by Nieman Lab. The update followed fierce backlash from the publishing industry over earlier language that appeared to let Meta use content from associated websites to develop, research and train its products.
The revised terms no longer say Meta will "train" its products on scraped content. They do, however, still permit the use of scraped content to develop any "business product," according to Nieman Lab.
Legal experts quoted by the outlet, including Cornell Law School professor James Grimmelmann and media and intellectual property attorney Matt Topic of Loevy & Loevy, said the changes do not prevent Meta from scraping entire sites or training AI on that content.
Meta spokesperson Daniel Roberts said content owners can use robots.txt to signal their preferences, and that the update does not grant rights superseding those preferences.
When the terms first appeared, legal experts told Nieman Lab they could let Meta argue its commercial terms supersede blocking instructions such as robots.txt. Meta then removed the training language after the publishing industry's objections.
Quick answers
When do Meta's revised business terms take effect?
The revised terms take effect October 30th.
Does Meta's updated policy still allow AI training on scraped content?
According to legal experts quoted by Nieman Lab, the revised terms still allow Meta to scrape sites and train AI on the content, though Meta spokesperson Daniel Roberts said robots.txt preferences are not superseded.