APDCAT turns AI rights assessments into a free desktop tool
Catalonia’s data protection authority APDCAT released a free application for assessing how AI systems affect fundamental rights. Organizations describe their system, select affected rights and rate potential harm. They can then enter safeguards and recalculate the remaining risk. The software puts a methodology introduced in 2025 into a guided workflow, with results intended as an indicative self-assessment.
Artificial Intelligence··Morning
Organizations rate the harm to each affected right
Catalonia’s data protection authority APDCAT released a free desktop application for assessing the impact of artificial intelligence on fundamental rights. The tool implements FRIA, the Fundamental Rights Impact Assessment method. Organizations and developers begin by describing the system’s objectives, characteristics, intended countries and data types. They then identify the rights that could be affected and can add rights beyond the supplied list.[1], [2]
Safeguards are entered before remaining risk is recalculated
For each right, the assessment considers probability, exposure, gravity and the effort needed to overcome harm. Each factor has four levels, from low to very high. Existing controls and affected groups are part of the assessment. Preliminary results help identify where measures are needed; after entering those measures, users can recalculate residual risk and export a PDF. APDCAT describes the results as an indicative self-assessment based on the organization’s inputs.[1]
The method predates the new application
The Catalan method was developed with expert Alessandro Mantelero and a community of data protection officers. It was presented to Catalonia’s Parliament in January 2025 and later tested on four AI prototypes at the Cruïlla festival. The application offers Catalan, Spanish and English interfaces on Windows, Mac and Linux, with example cases to guide its use.[1]
