Regulators in the UK and the European Union have compelled Meta to halt its plans to train AI on user data. The Irish Data Protection Commission (DPC) and the UK’s Information Commissioner’s Office (ICO) have opposed Meta’s decision to use public content shared by users on Facebook and Instagram for training large language models (LLMs).
Acting on behalf of several European data protection authorities (DPAs), the DPC requested Meta to delay the training. Meanwhile, the ICO asked the company to pause and reassess its plans.
Meta expressed disappointment with these requests, describing them as a setback for European innovation.
Both regulators welcomed the pause and committed to continued engagement with Meta and other generative AI developers to ensure user rights are upheld.
Relying on legitimate interest The regulators’ actions followed Meta’s recent announcement to begin training LLMs on user data. In response, the privacy rights group noyb (“none of your business”) filed 11 complaints with DPAs across the EU to pressure Ireland’s DPC.
Noyb argued that Meta’s plans violated “at least ten” GDPR articles, including the requirement for users to opt in to personal data processing.
In this case, users couldn’t opt out easily; they could only fill out an objection form hidden in Instagram’s settings, with Meta having the discretion to honor it.
Meta defended its practices by claiming “legitimate interest” as justification for data processing, an argument often rebuffed by European authorities.
On a webpage about the policy change, Meta stated that legitimate interest provides the appropriate balance for processing public data at the scale needed to train AI models while respecting user rights.
Noyb welcomed the regulators’ pushback. However, according to chair Max Schrems, they will continue to monitor the situation closely.
“So far, there has been no official change to the Meta privacy policy that would make this commitment legally binding. The cases we have filed are ongoing and will require an official decision.”