France’s data protection authority, the Commission nationale de l’informatique et des libertés, is partnering with fellow French digital regulators and research organizations to develop a catalog of tools that evaluate AI models’ ability to keep personal data confidential if it is trained with such information. The CNIL will oversee the effort, called PANAME, or the Privacy Auditing of AI Models, project with the hope of releasing tools that are either fully or partially open source.
Full story