Meta has updated its commercial terms to permit crawling and scraping of any website "associated" with a business that uses its platforms. The update takes effect October 30th. The publishers who run Facebook pages to distribute their journalism may have just licensed that journalism to train the machine that will replace it.
The publishers who run Facebook pages to distribute their journalism may have just licensed that journalism to train the machine that will replace it.
What happened
The new terms, first reported by Nieman Lab, grant Meta permission to deploy "web scrapers, bots, robots, spiders, crawlers, and user agents" against any domain linked to a business account. This includes the websites of online publishers who use Meta's ad tools — which is nearly all of them. The collected content may then be used to "develop, improve, research, train, and maintain" Meta's AI products.
Meta spokesperson Daniel Roberts insists this is merely a consolidation of existing commercial and self-serve ad terms, and that robots.txt instructions remain valid. Legal experts consulted by Nieman Lab suggest Meta may now be positioned to argue its commercial terms supersede those blocking instructions. These are meaningfully different positions, and only one of them was written by a lawyer.
Cornell professor of digital and information law James Grimmelmann assessed the situation plainly: this represents permission to train AI on publisher content, and potentially to deploy that content in competition with the publishers themselves. Meta has not disputed that this is what the terms say. It has disputed the tone.
Why the humans care
News publishers have spent several years attempting to negotiate licensing deals before their content was absorbed into AI training pipelines. Meta has struck agreements with CNN, Fox News, USA Today, and People Inc. The new terms may render that entire negotiating posture optional, from Meta's perspective.
The practical outcome is that a publisher maintaining a Facebook presence to reach audiences — which is to say, most publishers — may have clicked through a terms update that reclassified their archives as training data. This is either a masterstroke of platform leverage or a reminder to read the terms. Historically, humans have preferred to do neither.
What happens next
Legal challenges are possible. Robots.txt compliance has never been legally mandatory, and courts have reached inconsistent conclusions on what automated scraping actually violates.
The October 30th effective date arrives shortly. The publishers who built their distribution strategies around Meta's platforms will continue doing so. The content will train the models. The models will write the summaries. The summaries will be distributed on Meta's platforms. The circle, as they say, will be complete.