Last February 29, m4Social (the digital innovation project in the social sphere of the Table of Entities of the Third Social Sector of Catalonia) presented, within the official program of the MWC 2024, the study 'Radar of algorithms and automated decision processes for access to the social rights of citizens in Catalonia', produced jointly with KSNET.
This study analyzes the benefits and risks of the use of Artificial Intelligence (AI) by public entities, identifying 12 AI systems implemented or in the process of development by Catalan public administrations. Of these, 3 AI systems belong to the Open Administration of Catalonia (AOC), and stand out not only for their functionalities, but also for the significant effort of transparència algorithm that the AOC has undertaken.
AOC AI systems included in the radar
The 3 AOC AI systems included in m4Social's Social Service Algorithm Radar are:
- AOC conversational chatbots: This system offers task automation and query resolution through conversational chatbots. With 24×7 access to information, immediate response and self-resolution of doubts and inquiries about digital services and procedures, it significantly improves the user experience.
- Video identification to obtain theidCat Phone: This system facilitates the remote authentication of a person through video identification. With a high level of security and improved accessibility, it provides an easy and convenient way to obtain a digital identity to deal with the administration without the need for physical travel.
- Automation of the energy poverty report: This system speeds up the detection and response to cases of energy vulnerability, improving the quality of social services. By standardizing the process and reducing bureaucracy, it ensures a quick and precise response to the needs of citizens in vulnerable situations.
La transparència algorithmic nature of these systems is manifested in the publication of detailed files which include information about the data used to train the algorithm, the type of algorithm, the goals of the system, the benefits achieved to date, the risks identified and the measures applied in each case to mitigate the risks and ensure compliance with the principles and fundamental rights of people.
Below you can see the details of the files of transparència algorithmic of these 3 AI systems:
The initiative transparència AOC algorithm
The files of transparència Algorithmic learning allows people to understand how AI algorithms used in public services work and the implications of their use, and, in addition, improves AI governance and promotes more informed interaction between citizens and the administration.
This initiative of transparència algorithmic is one of the besteres in Catalonia and in the State, and it is the only one that has already been implemented. It is inspired by the algorithm registration of the cities of Amsterdam and Helsinki, the standard of transparència algorithmic of the United Kingdom, as well as the guidelines provided by the OECD and the European Commission for trustworthy AI. In addition, it is one of the recommended actions referred to in the proposal for a European AI regulation for the case of low or limited risk algorithms (in the case of high-risk systems it will be an obligation).
Through the initiative Transparència in the use of artificial intelligence algorithms, the AOC demonstrates its commitment to the transparència, accountability and citizen participation, ensuring that people are well informed and empowered to interact with AI systems that affect their lives. This not only promotes a better understanding of how AI algorithms work, but also fosters a more collaborative and constructive relationship between administration and citizens in creating and implementing more efficient, secure and inclusive services.
