Meta Removes Clause Raising Concerns Over Using Company Content for AI Training

Meta has removed a clause from its commercial terms of service that legal experts viewed as broad permission for the company to use content from its clients' websites to train AI models, as reported by TechCrunch via The Verge.

The company had updated its terms, stating that it could use any content from domains, web pages, or links associated with business accounts to use, develop, improve, research, train, and maintain its products. Experts speaking to Nieman Lab considered this a notable clause that could grant Meta the right to crawl publishers' websites to train its models.

However, a Meta spokesperson confirmed that the update did not reflect a change in how the company crawls the web. The company later announced the removal of the clause, stating that the crawling terms were limited to its advertising and commercial products and had been further clarified, emphasizing that this does not change how it uses web crawlers.

Observers believe the incident reflects growing sensitivity among publishers regarding the use of their content to train AI tools, especially as the company continues to face lawsuits and legal issues related to intellectual property rights and data usage.

What Do These Terms Mean?

Commercial Terms of Use: A contract governing the relationship between a digital platform and the clients using its commercial services.

Web Crawling: An automated process that browses websites and collects their pages for indexing or analysis.

Model Training: Feeding an artificial intelligence model large amounts of data to learn patterns and generate responses.