Overview
- Meta started feeding its Meta AI model with publicly available Facebook and Instagram content on May 27 after the opt-out deadline passed
- WhatsApp’s end-to-end encryption still protects private chats but any messages sent to Meta AI or in public groups can be used for training
- Users cannot fully disable Meta AI across WhatsApp, Instagram or Facebook and are urged to avoid sharing sensitive information with the chatbot
- EU regulators and consumer organizations have raised legal challenges over the use of personal data without explicit opt-in consent
- Meta argues that incorporating EU user content is crucial for tailoring its AI to regional languages and norms, though data already ingested cannot be removed