Meta’s upcoming changes to its commercial terms of service will compel business accounts to permit Meta’s automated systems to crawl and use their linked website content for AI training, a move that legal experts warn could undermine licensing agreements and impact news publishers globally.
- New Meta terms require business accounts to let AI training on linked web content.
- News publishers face potential loss of licensing fees and control over their content.
- Meta claims no change in crawler usage but allows no opt-out for businesses.
What happened
Meta has announced an update to its commercial terms of service that will take effect on October 30, requiring business accounts to authorize Meta’s automated software to access, index, crawl, and use content from websites linked to their accounts, not just social media posts. This expanded permission is aimed at allowing Meta to gather data to provide, improve, and train AI features across its platforms. The update consolidates previously scattered crawling policies into one streamlined document without explicitly mentioning AI but strongly implying its use for training.
The new terms do not permit businesses to opt out of this data collection. Instead, businesses are effectively told to cease using Meta services if they refuse to grant permission. Meta spokespersons maintain the company will respect robots.txt exclusions for news sites but indicate that the terms provide Meta with sweeping rights that could challenge existing licensing models and raise complex legal questions about the legality and scope of AI training on third-party content.
Why it matters
This change has sparked significant concern among news publishers and legal experts because it could undermine millions of dollars in licensing fees typically charged for news content used in training AI models. With ongoing litigation over whether large language models training on news content constitute fair use, Meta’s terms may provide the company with a broad legal foothold to scrape and use such content freely, potentially bypassing licensing payments entirely.
Experts warn these terms grant Meta unusually broad rights to use content, potentially allowing it to train AI on any material from linked websites and even deploy those outputs in competition with original publishers. For the media industry, which relies on licensing revenues, this shift could disrupt business models and diminish control over proprietary content. Moreover, Meta’s recent moves to monetize link sharing reinforce concerns that the company is leveraging commercial agreements to gain unrestricted access to valuable content.
What to watch next
The reaction of courts to Meta’s revised terms will be critical in determining their enforceability and the broader implications for AI training content rights. Legal challenges, particularly from news publishers, could shape how far Meta and other platforms can push automated content scraping without licenses. Any rulings favoring or limiting such use could set important precedents for digital content licensing and AI development strategies.
In parallel, monitoring how Meta enforces these terms—whether it honors robots.txt instructions or uses the update to push broader access—will be key. Businesses and publishers should review their use of Meta services closely and consider technical and legal protections. The unfolding policy and legal response will likely influence industry standards and regulatory attention on AI training transparency and fair content use.