Meta to Train AI Models Using Public U.K. Facebook and Instagram Posts

Meta has stated that it would begin training its artificial intelligence (AI) systems using public information published by adult users across Facebook and Instagram in the U.K. in the coming months.

"This means that our generative AI models will reflect British culture, history, and idiom, and that U.K. companies and institutions will be able to utilize the latest technology," the social media giant stated.

As part of the process, users aged 18 and above are anticipated to get in-app notifications beginning this week on both Facebook and Instagram, detailing its modus operandi and how they may simply access an objection form to prohibit their data being used to train the company's generative AI models.

The business stated it would uphold customers' decisions and that it won't approach consumers who have previously objected to their data being used for their purpose. It also highlighted that it would not contain private chats with friends and relatives, as well as information from accounts of children.

Furthermore, Meta stated the finding is the consequence of its interaction with the U.K. Information Commissioner's Office (ICO) and its guidelines supporting Meta's adoption of the legal foundation of Legitimate Interests, which it said is a viable method for utilizing first-party data to train its AI models.

"While our original approach was more transparent than our industry counterparts, we've incorporated feedback from the ICO to make our objection form even simpler, more prominent and easier to find," Meta continued.

It's worth mentioning that Meta has suspended comparable activities in the European Union after a request from the Irish Data Protection Commission (DPC) as of June 2024. It labeled the decision a "step backwards for European innovation."

Austrian privacy non-profit noyb has subsequently criticized the firm of shifting the responsibility on users – i.e., making it opt-out as opposed to opt-in – and failing to offer appropriate information on how the corporation is planned to utilize the publicly-accessible Facebook and Instagram data.

The news comes as Meta banned the usage of generative AI in Brazil after the country's data protection authorities issued a preliminary ban objecting to its new privacy policy.

The ICO, in reaction to Meta's intentions, said it wants to monitor the situation once the firm informs consumers and starts processing their data.

"We have been clear that any organization using its users' information to train generative AI models needs to be transparent about how people's data is being used," Stephen Almond, executive director of regulatory risk at the ICO, said.

"Organizations should put strong controls in place before they start utilizing personal data for model training, including offering a clear and straightforward method for consumers to object to the processing. The ICO has not issued regulatory permission for the processing and it is for Meta to guarantee and show continuous compliance."