Meta has pushed the start date of its new business terms from October 30 to November 20, after publishers raised concerns that the language could let the company scrape news sites for AI training.
The terms apply to news sites that use Meta's platforms. They appeared to give Meta permission to freely scrape that material and use it to train AI systems. Publishers read them that way and objected. Meta spokesperson Daniel Roberts said the new date is meant "to allow more time for us to explain to our partners and stakeholders what's changing and what's not." That buys 21 days.
The statement does not say the wording will change, and that is the detail worth watching. Roberts frames the problem as one of explanation: partners may have misread the terms, and Meta needs time to clarify. If that is right, the text stays and the messaging changes. If publishers read the language correctly, an extension only delays the argument.
The dispute is bigger than one contract. AI developers have generally leaned on the idea that publicly available web content is fair game for training, while publishers argue they should be asked and paid. Normally a publisher can try to block a crawler. Here the permission sits inside business terms attached to using a platform, so refusing it may mean giving up the platform itself. That makes a contract clause a sharper lever than a scraper.
A delay also leaves room for a quiet edit. Whether the version that takes effect on November 20 matches the one publishers objected to is the question Meta has not yet answered.