From The Verge · Emma Roth · · 1m 10s
Meta's terms let it train AI on news for free
New policy allows Meta to scrape publisher websites without licensing deals, bypassing agreements it struck with major outlets.
In brief
New policy allows Meta to scrape publisher websites without licensing deals, bypassing agreements it struck with major outlets. Meta's new terms effective October thirtieth allow it to crawl and scrape websites associated with business accounts, including news publishers, to train AI models without licensing fees…
Meta's new scraping rights
Meta's updated terms, taking effect October thirtieth, allow the company to crawl websites associated with business accounts, including news publishers, to train AI models.
The broad scraping clause
The terms authorize Meta to use automated software like web scrapers and crawlers to access, index, and collect content from domains associated with products or services.
“use automated software (such as web scrapers, bots, robots, spiders, crawlers, and user agents)”
Bypassing licensing deals
Meta has signed AI licensing agreements with CNN, Fox News, USA Today, and People Inc., but these new terms could let it access far broader content without paying.
The compliance pressure
Businesses that reject Meta's new policies risk losing access to Facebook, Instagram, WhatsApp, and Messenger for commercial purposes, according to the terms.
“may stop all access or use of our Products for business or commercial purposes”
The double charge
Last month Meta launched a subscription requiring businesses to pay to include links in posts and reels, while simultaneously gaining free access to scrape that same content.
“Meta has begun charging users to post links to the same content that it will be permitted to crawl for free”
The one thing
Meta's new terms effective October thirtieth allow it to crawl and scrape websites associated with business accounts, including news publishers, to train AI models without licensing fees.
