Meta delays new business terms after publisher pushback on AI scraping

News publishers can prevent Meta's AI from training on their content by using robots.txt.

Meta has postponed its new business terms, originally slated for October 30th and later November 20th, following significant pushback from publishers and advertisers. These publishers expressed concerns that the revised terms could permit Meta to scrape news sites for artificial intelligence training, despite Meta's claims of removing specific "training" language. The News/Media Alliance formally requested Meta CEO Mark Zuckerberg revise the terms, citing potential "severe harm." Digital advertisers and publishers in the EU and UK also lobbied regulators to halt the terms. Meta spokesperson Daniel Roberts stated the delay is intended to provide more time to clarify its web crawling practices and the continued availability of robots.txt for content owners to manage site access.

The delay indicates that Meta is sensitive to the intense scrutiny and potential regulatory action from publishers and advertisers regarding its AI data collection practices. Publishers' ability to block AI scraping via robots.txt, if effectively communicated and enforced, could set a precedent for how platforms access and utilize copyrighted content for AI development.

This situation highlights a growing conflict between AI developers needing vast datasets and content creators seeking to protect their intellectual property and revenue streams. The outcome of Meta's negotiations and Meta's subsequent actions will signal the broader industry's direction on fair use, licensing, and compensation for data used in AI training.