The French data protection authority, the Commission nationale de l’informatique et des libertés (CNIL), and the Institut national de recherche en sciences et technologies du numérique (Inria) have signed a fresh partnership agreement aimed at strengthening their scientific and technical collaboration.

With regard to personal data protection and privacy, the two institutions intend to deepen their research into methods that mitigate the risks associated with the use of sensitive data. They will also promote the development and dissemination of technologies that enhance protection, such as anonymization, pseudonymization, and federated learning.

Addressing the AI Act Requirements

On the evaluation of algorithms and artificial intelligence, the agreement envisions the creation of audit and certification methods aligned with the European AI Act. The plan includes developing tools to measure performance, robustness, and fairness of systems, with practical experiments planned in concrete sectors like health, cybersecurity, and digital public services.

There is also a strong emphasis on transparency and explainability of algorithms. The objective is to make how these systems operate more accessible to users and regulators. The work will address decision-tracing techniques for automated decisions and examine their potential impact on fundamental rights.

Finally, the agreement provides a dedicated component for training and knowledge dissemination. It calls for the production of shared educational resources to raise awareness among professionals, researchers, and the general public about the ethical and legal dimensions of AI. Other initiatives include organizing seminars and workshops and publishing scientific papers to share the outcomes of these investigations.

Since the General Data Protection Regulation (GDPR) has been in force since 2018 and the European regulation on artificial intelligence (AI Act) has entered into effect, CNIL’s missions have broadened, notably in supporting both public and private sector actors and in supervising AI systems. De facto, Inria plays a key role as the scientific lead for the National Strategy for Artificial Intelligence (SNIA) and contributes to the creation of INESIA, the National Institute for the Evaluation and Safety of Artificial Intelligence.