Meta has delayed the rollout of updated business terms after facing resistance from publishers. The change comes after media organizations expressed concern that the new language would grant the company permission to scrape news websites to train its artificial intelligence models.

According to a statement from company spokesperson Daniel Roberts, the new terms are now scheduled to take effect on November 20th. This postponement is intended to provide more time to address the issues raised by affected parties.

What it means

The conflict highlights the ongoing tension between AI developers and content creators over data acquisition. Publishers are increasingly wary of terms that allow large-scale automated harvesting of their intellectual property to fuel machine learning systems without explicit protections.