Meta Embarks on New AI Initiative Using Public Data in the UK
Meta, the parent company of Facebook and Instagram, has announced a significant move to enhance its artificial intelligence (AI) capabilities. In the coming months, Meta will commence training its AI models using publicly shared content from adult users in the United Kingdom.
Training Objectives
The company aims to integrate aspects of British culture, history, and idioms into its AI models. This initiative will enable UK businesses and institutions to leverage advanced technology resonant with local nuances. Meta emphasized the importance of culturally relevant AI for better local engagement and application.
User Notification and Consent
Starting this week, users aged 18 and above will receive in-app notifications on Facebook and Instagram. These notifications will outline the process and provide an easy-to-use objection form for those who wish to deny the use of their data for AI training.
User Privacy and Data Protection
Meta has assured users that it will respect their choices regarding data usage. Private messages with friends and family, as well as data from minor users, will not be included in this AI training initiative.
Regulatory Compliance
Meta’s approach has been developed in consultation with the UK’s Information Commissioner’s Office (ICO). However, it is essential to note that ICO has not given explicit regulatory approval. Instead, Meta’s use of the ‘Legitimate Interests’ legal basis has been highlighted as a mechanism to use first-party data for AI training.
Stephen Almond, the Executive Director of Regulatory Risk at ICO, stated, “Any organization using personal data for AI model training must ensure transparency and provide an effective opt-out mechanism for users.”
Feedback and Adjustments
In light of feedback from the ICO, Meta has made its objection form more accessible and easier to find, aiming for a more transparent user experience. This move follows a suspension of similar efforts in the European Union after the Irish Data Protection Commission requested a halt in June 2024, labeling it as a setback for European innovation.
Criticism and Controversy
Despite these efforts, Meta has faced criticism for its opt-out approach, as opposed to an opt-in system. Austrian privacy non-profit noyb has accused Meta of not providing adequate information on how publicly accessible data will be used.
Global Context
Recently, Meta also suspended the use of generative AI in Brazil due to objections raised by the country’s data protection authority. This global perspective showcases the varying regulatory landscapes Meta must navigate as it expands its AI training initiatives.
Future Implications
As Meta proceeds with its plans in the UK, the ICO will monitor the situation closely to ensure compliance and safeguard users’ data privacy. This development marks a significant step for AI innovation, balancing technological advancements with user trust and regulatory adherence.