Experts say Meta's new terms of use will allow scraping of websites, including news sites, "associated with" business accounts on its platforms, with no opt out

Meta can now collect content from your business website to train its AI, even if you don't post it directly on their platforms.

Starting October 30, Meta's updated commercial terms will permit the company to scrape websites associated with business accounts, including news organizations, to train its AI products. This change, revealed by legal experts who reviewed the terms, broadens Meta's data collection capabilities beyond content directly posted on its platforms. Business accounts that continue to use Facebook, Instagram, or WhatsApp will be deemed to have consented to these terms, with no explicit opt-out available. Users are told that if they do not agree, they must cease all business or commercial use of Meta's products. Meta stated the update consolidates existing crawling language into one place and suggested content owners can still manage access via robots.txt, though legal experts believe the new terms could supersede such instructions. This move comes as Meta faces ongoing lawsuits regarding AI training data and follows similar bundling strategies seen with other tech giants.

This policy change fundamentally alters how Meta can leverage external web content for AI development, potentially shifting the landscape for content licensing and data ownership. By requiring businesses to consent to scraping as a condition of using their services, Meta is effectively creating a broad, low-cost data feedstock for its AI initiatives. This tactic mirrors strategies employed by other large tech companies, suggesting a wider industry trend towards consolidating control over data necessary for advanced AI model training. The lack of a clear opt-out mechanism forces businesses into a difficult choice: either accept Meta's data harvesting or forfeit access to critical platforms for customer engagement.

The implications for news publishers are particularly significant, as their websites often contain valuable, curated content that is expensive to produce. Meta's ability to scrape and train on this content without direct compensation or explicit consent could devalue their intellectual property and potentially enable Meta to compete more directly with them using AI-generated insights derived from that content. The move also raises concerns about the enforceability of existing webmaster tools like robots.txt, as Meta's terms may assert precedence, creating a new legal battleground over digital content rights and AI development.