Meta’s new commercial terms say if its AI makes a mistake, it will legally be the customer’s fault

Meta plans to update its commercial terms on October 30, shifting responsibility for errors, inaccurate outputs, or unsuitable results from its AI features to enterprise customers. The proposed terms state that customers are solely responsible for all risks associated with Meta AI features and resulting content, while also extending responsibility to content used or uploaded by agencies, service providers, partners, and contractors acting on a customer’s behalf. The proposed terms grant Meta a license to use intellectual-property-protected content made available in connection with Meta products to provide, develop, improve, research, and maintain its products. They also authorize automated tools including crawlers, bots, and web scrapers to access, index, cache, analyze, collect, and retrieve content from domains, webpages, and URLs associated with customers’ products or services. Meta said the crawling language applies only to its advertising and business products, does not alter its use of web crawlers, and does not override content owners’ robots.txt preferences; a Meta spokesperson said wording changes were forthcoming. Meta’s earlier July changes made certain licenses perpetual and stated that they survive termination of the terms. Analysts and consultants cited broad content licensing, automated crawling, and AI-risk allocation as major concerns, while Meta may still revise the proposed wording before the October 30 effective date.