Should using Facebook and Instagram give Meta permission to use data from a business’ website to train AI? That’s the question behind a recently proposed terms of service update that, after creator backlash, has now been delayed.
Earlier this month, Meta previewed an update to the platform’s terms of service that some experts argued would give the company permission to scrape data from business websites to train its AI platform.
The updated terms go beyond Meta’s current permissions for using Meta platforms and add terms that allow Meta to use automated software such as crawlers and bots to analyze the business’s website.
The proposed terms would allow for using the business website, not just content posted to Instagram and Facebook, prompting concern across a number of groups from news publishers to photographers and creators.
The proposed terms update initially would give Meta permission to “use any content from such domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of, the Meta Company Products, services, and features.”
The update initiated a rash of new videos and posts from creators concerned that the terms would allow Meta to use their work to train its AI system, including a US Senator who worried the update could harm businesses and creators.
After the creator backlash, that update has now been delayed from October 30 to November 20. Meta has also now removed the word “train” from the list of permissions actions it could take with a business’s website content. Meta has since changed the language from using that data for “Meta Company Products” to “advertising and business products.”
The proposed terms update is for public business accounts, not private accounts, but would affect public profiles and businesses, including photographers with public profiles on Instagram or Facebook.
The News/Media Alliance also sent a letter to Meta CEO Mark Zuckerberg demanding the withdrawal of the updated commercial terms, as publishers expressed concern that the terms' language would allow Meta AI to use content from news websites for training without licensing.
Much of the backlash from the originally proposed terms update comes from the lack of an option to opt out.
In a statement, Meta said that it delayed the terms “to allow more time for us to explain to our partners and stakeholders what’s changing and what’s not, including how Meta uses web crawlers and the fact that content owners will still be able to use robots.txt to signal preferences.”
Robot.txt files can serve as signals to AI platforms to block training on web platforms that support adding those files to the root directory, though researchers note crawlers are not required to follow those restrictions.
You may also like
Take a look at the best photo editing software or the best cameras for professional photographers.
