AI News · The messy stuff ·
Meta removes terms that could have let it train AI on business websites
Meta has removed a provision from its updated commercial terms that said it could use content from a business account's domains, web pages or URLs to develop and train its products. Legal experts had warned the language could let Meta scrape publishers' sites for AI training.
Key points
- Meta's updated commercial terms covered content from business account domains and URLs.
- Legal experts told Nieman Lab it could allow scraping of publishers' sites for AI training.
- Meta removed the provision and said it does not change how it uses web crawlers.
What happened: Meta recently updated its commercial terms of use with language saying it could use content from a business account's domains, web pages or URLs to provide, develop, improve, research, train and maintain the integrity of its products and services, TechCrunch reported. The Verge has since reported that Meta removed the provision.
The details: Legal experts interviewed by Nieman Lab called it a remarkable provision and said it could let Meta scrape the websites of any publisher with a Meta business account and use that content to train its AI models. A Meta spokesperson said the updated terms did not reflect a change in how Meta crawls the web. In a statement, Meta said the crawling terms it consolidated applied only to its ads and business products, that it has clarified this in the terms, and that the change does not change how Meta uses web crawlers.
Background: Meta won a high profile copyright lawsuit over AI training on books last year, but other cases against AI companies, including Meta, are ongoing, TechCrunch noted. Questions over what content AI developers may collect and train on remain a central legal issue for the industry.
Who it affects: Publishers and other businesses with Meta business accounts were the group most exposed by the original wording. Meta has not disclosed how many accounts were covered by the terms before the change.
Our take
Teams that sign up for vendor business accounts should read updated terms closely, since data use clauses can quietly expand what a provider may train on.