On May 16, 2023, the French Data Protection Agency (“CNIL”) published its Action Plan on Artificial Intelligence (“AI Action Plan”). The AI Action Plan builds on his CNIL’s previous work in the AI field and comprises a series of activities undertaken by the CNIL to support the deployment of AI systems that respect individual privacy.
The AI Action Plan is centered around four main goals:
- Understanding the capabilities of AI systems and their impact on people: Through new artificial intelligence services, CNIL will focus on solving key data protection issues related to the design and operation of AI applications. These issues include protecting data published on the web from scraping, protecting data submitted by users of AI systems, and protecting data collected for training AI applications and output generated by AI. This includes the impact on rights to personal data regarding. Problems such as AI systems.
- Guidance on AI development that respects personal data: To support organizations innovating in the AI field and prepare for the possible passage of EU AI law, the CNIL has issued guidance and guidance on several AI topics, including a guide on the rules governing the sharing and reuse of AI. Publish best practices. It not only provides data, but also recommendations for designing generative AI systems.
- Supporting innovative players in the French and European AI ecosystem: CNIL supports actors in the AI ecosystem who innovate in ways that ensure the protection of fundamental rights and freedoms in France and Europe. In particular, the CNIL plans to soon invite new projects to participate in the 2023 regulatory sandbox and increase dialogue with research teams, R&D centers, and organizations developing AI systems.
- AI system audit and control: The CNIL will develop tools to audit AI systems and will continue to investigate complaints filed with the agency related to AI, including generative AI.
