Revised terms of service published by Meta now include provisions allowing the company to draw on content from business users' websites to develop its AI products, with no payment or formal agreement required from those users. Scheduled to take effect on Oct. 30, the policy covers all entities holding free business accounts — a scope that attorneys and media professionals say extends to news organisations, film studios, and recording artists, among other affected parties.
Under the updated terms, Meta would be permitted to deploy automated tools to collect content from the platform for use in building its own products, AI development among them. Under that reading of its own terms and conditions, Meta would gain authorisation to index an entire news outlet's website — and draw on that material to develop and maintain its range of products — simply because the publisher posted one article link on its Facebook page.
Businesses that object to Meta's terms of service put their continued access to the platform at risk. In a written statement, News/Media Alliance President and CEO Danielle Coffey argued that the policy's reach is broad enough to put the copyrighted online content of any business holding both intellectual property and a Meta account at risk of being taken without consent. Coffey also noted that copyrighted content hosted on a celebrity's own website could fall within Meta's reach, provided that person holds a business account on the platform.
A law professor at Emory University told a television outlet that Meta wields a degree of power over the broader media environment that even well-resourced publishing organisations are not insulated from. The attorney representing The Mercury News and seven fellow MediaNews Group outlets in their lawsuit against Microsoft and OpenAI characterised Meta's policy change as an attempt to either deceive users or push them toward giving up their content without receiving payment, describing Meta's conduct as lacking transparency, propriety, and integrity.
A Chicago-based media attorney drew a parallel between Meta's conduct and opportunistic land seizure, contending that AI companies — Meta included — are pulling substantial quantities of news content into their platforms to construct products that hollow out the revenue structures journalism depends on.
In an unsigned corporate statement, Meta maintained that its practices around web crawlers remain unchanged, characterising any assertion to the contrary as inaccurate, while leaving specific questions put to the company without a direct answer. According to the firm's statement, the move brings its commercial and self-serve advertising terms under a unified framework, with the stated aim of giving businesses a clearer view of how the platform's ad policies apply to them.
A social media post from Andy Stone, who holds the vice president of communications role at the company, conveyed that web crawler practices had not been altered. The purpose of the update, he said, lies in giving Meta a reliable way to pull product details from third-party retail sites when generating ad and content recommendations within its own platforms — a function he described as distinct from any harvesting of news content.
In an interview with a television news outlet, Eric Goldman — a law professor specialising in technology at Santa Clara University — characterised the wording in Meta's updated policy as unusually sweeping in scope. Goldman put forward the view that any organisation with truly limited goals would have written its revised terms in plainer, more accessible language. He also argued that Meta's contract wording left room for adjustment — for instance, by adding a robots.txt directive barring data collection and model training. Goldman further noted that, without a demonstrable commitment to good faith, Meta could theoretically invoke that clause as a mechanism to sidestep the robots.txt directive.
Publishers and other media outlets have grown increasingly cautious toward technology and artificial intelligence companies over concerns that proprietary content is being harvested or repurposed to train AI systems without authorisation. Deals struck between news outlets and technology firms now allow journalistic content to be adapted for other uses in return for yearly fees that can reach into the millions. Both a prominent television news network and Axios entered into content-licensing arrangements with Google.
A number of publishers — MediaNews Group among them — have taken these companies to court, with those legal disputes still under way.
