Archive — US — techcrunch
Meta to start training its AI models on public content in the EU
Meta plans to resume training its AI models on public content from users in the EU, reversing a previous pause due to regulatory concerns over data privacy. This follows discussions with regulatory bodies to ensure compliance with GDPR and aims to enhance AI understanding of European nuances. Notifications will be sent to users regarding data use and opt-out options are available.
Key points
- Meta's AI training plans were paused due to EU regulatory pressures related to GDPR.
- The company is now set to resume AI training using public user data in the EU.
- Users will receive notifications about data usage and can opt-out if desired.
- Meta's strategy seeks to make AI more culturally relevant to European users.
- EDPB's opinion helped clarify legal obligations for Meta, allowing training to proceed.
Why it matters
- Meta's actions reflect broader challenges tech firms face in balancing AI advancements with data protection laws.
- Similar moves by Google and OpenAI illustrate ongoing trends in utilizing public content for AI development.
- The scrutiny from regulators like the DPC indicates a shifting landscape for AI training in Europe.
Meta's decision to train AI on EU public data demonstrates a significant step towards compliance with privacy laws, highlighting the importance of regulatory dialogue while enhancing user engagement with AI tailored to their cultural context.