Revised Meta terms may still permit news site scraping, experts say
Meta has removed explicit wording from its business terms that said it could use website content to train its AI, after publishers objected. Legal experts say the revision may still leave room for Meta to use scraped material in AI development, so the dispute over what the terms permit remains unresolved.
The terms take effect on 30 October and apply to websites associated with business accounts. Cornell law professor James Grimmelmann says the wording still does not clearly restrict scraping entire sites, while attorney Matt Topic argues that using content to “develop” a business product could include training AI. Meta says the update does not grant new rights over robots.txt preferences and is not intended to enable free scraping of news sites.
In short
- Meta removed explicit AI training language from its business terms.
- Legal experts say broader wording may still cover AI development.
- Meta says robots.txt preferences remain in place.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Meta’s removal of the explicit AI-training language may address publishers’ concern that the terms claimed a broad new right to their material. The company says the revision does not override robots.txt preferences and is not intended to enable free scraping, which could support its view that the terms govern business services without authorising unrestricted use of news content.
The case against
The revised terms may still leave publishers uncertain about how their material can be used, because they do not clearly rule out scraping whole websites. If “develop” a business product includes training AI, the wording could permit uses that publishers object to even without an explicit training clause, making clearer limits important for protecting their content.
Read the full article at the source →
Originally published by The Verge as “Meta’s AI could still scrape news sites, legal experts say”.