Meta has initiated a quiet but consequential update to its terms of service, sparking immediate pushback from news organizations and legal experts concerned about the company’s ability to scrape online content for artificial intelligence development. The changes, which are set to take effect on October 30, have triggered a firestorm of criticism regarding how the tech giant interacts with the intellectual property of digital publishers.
The Scope of the Revision
The controversy began when an investigation by Nieman Lab revealed that the original language in Meta’s updated terms appeared to grant the company broad rights to ingest content from sites associated with business accounts—which include virtually every major newsroom—to train its AI models. Following that report, Meta revised the language to specify that the content can only be utilized for “advertising and business products” rather than the entirety of its product suite.
Despite this narrowing, the update offers no path for business account holders to opt out before the October 30 effective date. Furthermore, Meta has notably omitted any explicit, binding pledge to honor the “robots.txt” files that websites use to signal which parts of their digital property should remain off-limits to automated crawlers.
Legal Experts Raise Alarm on AI Training
While a Meta spokesperson, Daniel Roberts, maintained that the changes are intended to provide clarity rather than acquire new scraping rights, legal scholars are unconvinced. The primary concern is that the term “business products” remains nebulous enough to encompass the development of AI.
Matt Topic, an intellectual property attorney at Loevy & Loevy, noted that the term “develop” is sufficiently broad to include the training of AI models. His assessment is shared by other academics, such as Cornell Law School professor James Grimmelmann, who argued that the revised language does not sufficiently prevent Meta from crawling entire websites. Additionally, Benjamin Sobel, a law professor at the University of Wisconsin, cautioned that without a legally enforceable guarantee, Meta’s adherence to robots.txt remains strictly voluntary and could be ignored at the company’s discretion.
Industry Backlash and Regulatory Scrutiny
The legal community is not the only group sounding the alarm. Prominent industry advocacy organizations have labeled the move as an aggressive overreach. The News/Media Alliance, led by president and CEO Danielle Coffey, described the terms as “outrageous and probably illegal,” arguing that they should not be permitted to take effect at the end of the month.
International pressure is also mounting. The European Publishers Council has announced plans to lobby E.U. regulators and the European Commission to perform a formal inquiry into the terms. Simultaneously, the U.K. News Media Association is calling for government intervention to protect the rights of publishers against what they perceive as a unilateral seizure of copyrighted material.
This friction occurs against the backdrop of a larger, ongoing legal battle between Meta and various content creators. In May, a group of publishers and author Scott Turow filed a class-action lawsuit against the tech giant, alleging that Meta used millions of copyrighted books and articles without authorization to train its Llama AI models. Meta has consistently denied these allegations and continues to mount a defense against the claims.
As October 30 approaches, the silence on opt-out mechanisms and the refusal to legally codify the status of robots.txt have placed Meta at a crossroads. For news organizations already wary of how their content fuels the current AI boom, these terms represent a significant escalation in the struggle to control digital assets in the age of generative technology. Whether regulators will step in to halt the implementation of these terms remains the most pressing question for the industry.
Disclaimer: This content is auto-generated for informational purposes only.
Source: Read Original News
