Meta is updating its website terms once again after facing backlash, with the social media giant now removing language that suggests it can scrape content from sites, including news websites, to train its artificial intelligence (AI) models.
The update, which will go into effect on 30 October, will take into account news publishers along with business accounts. This will affect all websites, especially news websites, which have at least one Meta business account.
The change which was published on 17 October removes a sentence that explicitly authorised Meta to use content collected from specified domains, webpages and URLs to “train” its products and services. The revised provision instead ties the stated purposes of crawling to advertising and business products and the integrity and security of Meta’s products.
But the revision does not include the underlying authorisation for automatic crawling. As per the updated Meta terms, bots, spiders, crawlers and other automated software can access, index, analyse, collect, and retrieve website content, as well as return periodically to collect updates.
What changed between the original and revised terms?
The original provision allowed Meta or entities under the company to use automated software to access, index, cache, analyse, collect, retrieve and crawl websites associated with a business’s products or services, which also included websites which are made available through the use of Meta’s platforms.
A separate sentence then stated that Meta could use the collected content to “provide, develop, improve, research, train, and maintain the integrity of” Meta products, services and features covered by the commercial terms and other applicable supplemental terms.
This is the sentence that triggered the controversy. that this could be interpreted as permission to collect material from the websites of the publishers for training AI models.
The revised working – In the revised English version, the sentence explicitly referring to training has been removed. The provision for crawling now connects the collection of contents from websites to the development, improvement and research of advertising and business products, besides maintainting Meta products’ security.
The authorisation to crawl websites and collect content remains.
Thus there are two crucial changes: it removes the explicit reference to training and narrows the stated purposes to advertising and business products rather than the broader category of Meta products and services.
It does not, however, explicitly restrict the use of collected information in every AI-related application, as per a comparison of the two versions by .
What rights do website owners retain?
As per Meta, the owners of websites can use a robots.txt file to communicate their preferences about automated access. However, this is a technical mechanism and not a legal enforcement tool.
Meta has also said that businesses can stop using its products for commercial purposes if they reject the new terms.
Meta has said that the new terms do not change how it uses web crawlers, and that owners of websites can continue using robots.txt to signal their preferences, Nieman Lab said.