Meta Scrubs Controversial Web Scraping Terms from Business Policy

Meta has removed a specific clause from its commercial terms of use that previously asserted the right to utilize content from external domains to develop and train Meta products. The policy shift follows widespread criticism regarding the breadth of the language, which legal observers noted appeared to grant the social media giant unfettered access to scrape publisher websites connected to Meta business accounts.

Scrutiny of data practices

Legal experts consulted by Nieman Lab identified the provision as a significant expansion of authority, suggesting it could allow the company to ingest protected editorial content for AI model development. While the company maintains that these terms were intended only to clarify operations for its advertising and business platforms, the removal suggests a pivot in response to mounting pressure surrounding how artificial intelligence systems ingest intellectual property.

In reporting reviewed by The Scale Report, a spokesperson for the technology firm emphasized that the update does not signal a change in the company's web crawling behavior. The revised terms now explicitly state that the consolidated crawling language applies strictly to business and ad-focused services, aiming to quell concerns that the policy functioned as a back door for general-purpose model training.

Context of the AI training landscape

The sensitivity surrounding this update is underscored by the current legal environment, where AI developers face numerous lawsuits over copyright infringement. Although Meta recently secured a victory in a separate high-profile legal challenge, other litigation continues to move through the courts. This friction reflects a broader tension between publishers seeking to protect their traffic and digital infrastructure and tech giants that rely on massive datasets to fuel their generative models.