Meta Agrees to Pause Plans to Train AI with User Data
Meta has announced that it will temporarily halt its efforts to utilize users' personal data for the training of artificial intelligence. This decision comes in response to concerns raised by the Data Protection Commission (DPC).
Data Protection Commission's Stance
The DPC has expressed its approval of Meta's choice to refrain from using public content from adults on its platforms, including Facebook and Instagram, to train its large language model within the European Union and European Economic Area.
Engagement between Meta and the DPC
Following extensive discussions between Meta and the DPC, the latter has confirmed that it will continue its collaboration with the company on this matter, in conjunction with other EU data protection authorities.
Meta, in a statement, expressed disappointment at the regulator's request to postpone the training of its LLMs with data from Facebook and Instagram. The company emphasized its commitment to compliance with European laws and regulations, highlighting the importance of transparency in comparison to other industry players.
Meta stressed the significance of incorporating local information into AI training to enhance user experience, noting that without such data, launching Meta AI in Europe would be challenging.
Future Collaboration and Adjustments
Despite the pause in training, Meta affirmed its ongoing cooperation with the DPC, citing that this delay would enable the group to address specific demands from the UK regulator, the Information Commissioner's Office, before proceeding with the training process.
Advocacy Group's Concerns
An advocacy group, NYOB, has called on regulators to take immediate action against Meta. The group raised alarms regarding Meta's updated privacy policy, which is set to allow the utilization of extensive amounts of user-generated content for AI training purposes.
(Pic: Artur Widak/NurPhoto via Getty Images)
