- Innovation
The Generalitat presents an Artificial Intelligence Guide to guarantee trust, transparency and responsibility
The heads of innovation services at the AOC, Carme Callés i David Pastor, have participated in the II Training Conference on Artificial Intelligence (AI) and Personal Data Protection, organized by the Catalan Data Protection Authority (APDCAT) with the collaboration of the School of Public Administration of Catalonia (EAPC).
The session dedicated to “AI in the local world” was celebrated on11 November, within the third day of training, with the aim of sharing good practices for deploying artificial intelligence useful, ethical and with guarantees in local public services.
In her speech, Carme Callés opened the session by presenting the AOC's commitment to responsible and transparent use of AIHe explained the algorithmic transparency reporting methodology, designed to help citizens understand how AI algorithms work, what their purpose is and how they can affect them, as well as to encourage participation, ensure standardized and comparable transparency and comply with legal transparency obligations.
He also explained how, during the preparation of each report, es manage risks to fundamental rights and ethical principles, and how this work is translates into a clear and understandable document that describes the measures applied to minimize risks. His speech concluded with the example of thealgorithmic transparency report partly AOC's generative AI chatbots.
For his part, David Pastor delved into the practical criteria to guarantee privacy, security and data protection in the deployment of AI systems, and presented two Featured use cases:
- Automation of energy poverty reports (IRER), a system based on programmed rules —not AI— but with potential impact on social rights, which already has its algorithmic transparency report published on the Transparency Portal.
- Anonymization of municipal plenary sessions, a minimal risk AI system that transforms documents to avoid disclosing personal data. It is currently in a pilot phase with 15 municipalities, and has already begun work on the algorithmic transparency reporting methodology, an iterative and collaborative process that involves all the actors involved in the design, development and implementation of the service.
Since June 2025, the AOC has taken significant steps in the field of ethics and algorithmic transparency:
- Creation of the Ethical and Responsible AI Working Group, within the community DPD Online (joint initiative APDCAT + AOC), to define a public standard for trust in AI that integrates the Catalan FRIA model (AI impact assessment on fundamental rights) and the practical transparency of algorithmic reports aimed at citizens. The group was launched on October 30.
- Integration of the AOC algorithmic transparency sheets at the Registration of AI systems and other algorithms of the Generalitat, to offer a global and coherent vision of the uses of AI in the Catalan public sector.
These advances reinforce the AOC's commitment to people-centered AI, which makes innovation, the protection of rights and citizen trust in digital administration compatible.
👉 You can check the presentation (PDF) and the video of the session (starts at 1:54:16).