Skip to content
K = 0.731Subscribe
Original
IndustryArs Technica · AIRelayed· Ashley Belanger·· 10 days ago

Meta drops the word "train" from its commercial TOS after AI backlash

Image: Ars Technica · AI
Who
Meta
What
removed language in its commercial terms that suggested it could use website content to train its AI
When

What happened

Meta removed the word "train" from its commercial TOS following AI backlash. Legal experts suggested that omitting explicit commitments to respect robots.txt could leave Meta able to scrape sites despite their stated preferences. Publishing groups in the UK, Europe, and the US are urging regulatory intervention, with the News/Media Alliance calling for the terms to be blocked before October 30.

Read next

See the full storyMeta Revises Business Terms After Publisher Backlash, Says Further Delay Is Possible

Reported by 3 sources

    The story so far

    Meta revised and postponed its proposed business terms after publishers challenged their implications for AI scraping, and now says implementation could slip beyond November 20 while consultations continue. Meta communications head Andy Stone said existing terms will remain in place until the update takes effect. The company had already moved the start date from October 30 to November 20. The dispute concerns permissions to use content from websites associated with business accounts promoting content across Facebook, Instagram, WhatsApp, and Messenger. The Verge initially reported that Meta had removed the controversial language entirely; Ars Technica described the revision as dropping the word “train.” Legal experts told Nieman Lab that the remaining wording could still authorize AI training: attorney Matt Topic said permission to “develop” business products could encompass training. Other experts said the absence of an explicit commitment to respect robots.txt could leave room for scraping despite website owners’ blocking instructions. Meta spokesperson Daniel Roberts said the update applies only to ads and business products, leaves crawling practices unchanged, and grants no new rights to override owners’ preferences expressed through robots.txt. Publishing groups in the UK, Europe, and the US have urged regulatory intervention, with the News/Media Alliance calling for the terms to be blocked.

    Meta / LlamaIndustry newsPolicy and regulation#Industry news#Policy and regulation#Meta

    Keep up with AI, one email a day

    Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue · 2026-10-31