Meta's New Commercial Terms Will Allow Company to Train AI on Business Website Content Without Opt-Out
Starting October 30, Meta's updated terms will require business accounts to permit the company to scrape and use linked website content for AI training, with no option to decline.
What happened
Meta is updating its commercial terms of service effective October 30, which will authorize the company to use automated scraping tools to access, index, and crawl content from domains and web pages linked to business accounts. The updated terms state Meta may use any such content to "provide, develop, improve, research, train, and maintain" its products and services. Businesses cannot opt out of this provision; the terms suggest they must either accept the update or stop using Meta services. The language refers to automated crawling and the word "train," though it does not explicitly mention artificial intelligence. Legal experts are raising alarms that this could allow Meta to avoid paying licensing fees that news organizations might otherwise charge for AI training access to their articles.
Context
This update is particularly significant for news publishers and media organizations, virtually all of which maintain Meta business accounts. News organizations are currently engaged in litigation with OpenAI over whether using copyrighted articles to train AI systems constitutes fair use or requires licensing fees. Meta's new terms appear designed to grant the company broad permissions regardless of how such cases are decided. The move would affect all types of business accounts, but the financial implications are most substantial for news publishers who could potentially charge millions in licensing fees for AI training access to their content.