The Higher Regional Court of Cologne concluded that Meta's processing of personal data for AI training purposes is permissible under the "legitimate interest" basis in Article 6(1)(f) GDPR, subject… | Luis Alberto Montezuma

The Higher Regional Court of Cologne concluded that Meta's processing of personal data for AI training purposes is permissible under the "legitimate interest" basis in Article 6(1)(f) GDPR, subject to certain conditions. The court found that Meta’s interest in developing and improving generative AI systems—such as chat assistants and content generation tools—constitutes a legitimate interest under GDPR. This includes economic interests, which are explicitly recognized as valid. Meta’s aim to develop AI tools that respond in real time and generate content was considered a real, present, and non-speculative interest. The court accepted that training large AI models requires vast amounts of real-world data. Alternatives like synthetic or anonymized data were found to be insufficient or impractical for achieving comparable results. Meta implemented safeguards like De-identification and opt-out mechanisms. The processing was not aimed at profiling or identifying individuals. --- CNIL's recommendations on the use of legitimate interest for AI Systems and models: https://lnkd.in/dF8rSdsa. CNIL's response to the consultation on the legal basis for AI development: https://lnkd.in/d6nrUAxK. CNIL's recommendations regarding conditions for implementing measures to be taken in the event of data collection by web scraping: https://lnkd.in/dcgjUpzh. ICO's consultation series on generative AI and data protection - The lawful basis for web scraping to train generative AI models: https://lnkd.in/dTiRpB2R. LfDI Baden-Württemberg's discussion paper on Legal basis in data protection in the use of artificial intelligence: https://lnkd.in/eW3inpFp. Dutch DPA's consultation on GDPR preconditions for generative AI: https://lnkd.in/dvm98SxM. German DPAs' Guidelines on Recommended Technical and Organizational Measures for the Development and Operation of AI Systems (many thanks to Carolin Loy): https://lnkd.in/dXCkrvZq. Estonian DPA's brief guidance on the legitimate interest legal basis: https://lnkd.in/dXPs7dGH. EDPB's opinion on AI Models: https://lnkd.in/d5ZxN6y2.