Meta Removes AI Training Policy After Website Data Controversy

Meta has faced scrutiny over controversial wording in its commercial terms that appeared to permit the company to use content from websites linked to business accounts for product development, research, security, and AI model training. Following criticism, Meta removed the disputed language from its terms.

In this TechWorld news report, we examine what the policy said, why legal experts and publishing observers raised concerns, and what Meta's response reveals about the ongoing debate over AI training data and copyright.

In this report, we explore:

• What Meta's disputed commercial policy language covered
• Why publishers and legal experts raised concerns about website content and AI training
• Whether linking a website to a business account constituted consent under the disputed wording
• What Meta said about removing the language and its web-crawling practices
• How web crawlers collect and process publicly accessible online content
• Why AI companies face continuing copyright disputes over training data
• What this controversy could mean for publishers, businesses, and independent content creators

Important context: The disputed wording raised concerns about the potential scope of Meta's rights to use website content. It does not, by itself, establish that Meta scraped every linked website, used all such content to train its AI models, or violated copyright law. Meta also stated that removing the wording did not change its internet data-collection methods. The claim that the revision was specifically intended to reduce legal liability remains an interpretation, not a confirmed explanation of the company's motives.