Creation of the AI+ research group

CoboiLab has launched the AI+ research group, a dedicated space for exploring how to integrate artificial intelligence (AI) into public services and innovation processes in an ethical, practical, and safe manner. This initiative arises from the growing presence of AI across multiple areas and the need to approach its use from a responsible perspective, which takes into account both the benefits and potential risks.

The creation of the AI+ group is the result of the collaboration between CoboiLab and the Department of Digital Transformation of the Sant Boi de Llobregat City Council. Its main goal is to develop tools and methodologies that ensure the application of AI aligns with values of transparency, equity and security. This involves creating an AI implementation protocol that includes a risk checklist and setting clear conditions for its public procurement. In addition, it aims to ensure that any AI application that is integrated complies with the ethical standards promoted by the European Union through the Artificial Intelligence Act.

During the working sessions held so far, the AI+ research group has extracted several important lessons that mark its roadmap:

1. The importance of prior planning: Initial approaches to AI focused on testing popular tools without conducting a thorough analysis or defining concrete objectives. This premature approach revealed the need to begin with a well-structured, methodological reflection before implementing new technologies.

2. European regulatory framework as a reference: The draft European AI Regulation has become a valuable resource for identifying risk levels and applying rigorous criteria in the design and implementation of AI applications. This regulation classifies AI systems into four categories:

   • Unacceptable risk: Mass surveillance systems without consent or social credits that classify people.

   • High risk: Personnel selection or biometric identification systems, which require strict security and transparency measures.

   • Limited risk: Applications such as customer service chatbots, which require adequate control but pose no major risks.

   • Minimal risk: Spam filters or personalized recommendations, with low impact on security or people’s rights.

3. Creation of an AI implementation protocol: The group is developing a protocol that includes a risk checklist and a clear operational framework to ensure that any AI project is deployed with ethical and social guarantees. This protocol aims to guarantee transparency, security and fairness in the use of AI within public administration.

The work of this research group is a first step towards building an innovation model that integrates AI in a conscious and responsible way. With a critical perspective and a rigorous methodology, CoboiLab aims to provide solutions that are not only effective, but also respond to the ethical challenges that this technology poses.