CiTIUS leads national discussion on Ethical Artificial Intelligence

The centre promotes a meeting at the Ramón Areces Foundation (Madrid) with national and international experts to reflect on the implications of technology in the current context, focusing on the responsible use of Artificial Intelligence (AI) and how to quantify its impact on everyday life.

The Ramón Areces Foundation, an institution known for its strong support of scientific research as a driver of progress and modernity, hosted the 'Social Impact of AI' conference last Thursday and Friday. The event aimed to initiate a serious debate on the humanistic values and ethical principles that should guide the development of Artificial Intelligence.

 The event was proposed by CiTIUS (Singular Research Center in Intelligent Technologies of USC), which brought together national and international experts to discuss the social impact of Artificial Intelligence during a forum organized by the center's researcher José María Alonso, coordinator of the National Network on Reliable Artificial Intelligence, an alliance of 12 research groups of recognized prestige in the area, known by the acronym 'Technological, Ethical, Legal, Social, Economic, and Cultural issues for Trustworthy AI' (TELSEC4TAI). 

TELSEC4TAI played a significant role in the conference, introducing from various perspectives the details of its common strategy to create social awareness about the importance of advancing towards a new generation of intelligent and 'responsible' devices. A series of new systems and technologies designed to respect and promote the principles of non-discrimination and equity, and which should also be capable of explaining their decisions in a comprehensible manner to the public. TELSEC4TAI combines the expertise of experts from various disciplines, with the aim of researching and developing technologies that promote security, social cohesion, and transparency in the use of Artificial Intelligence.

 Additionally, the meeting also served to present examples of the use of the protocol defined by Z-Inspection®, the first independent tool designed to assess the reliability and ethics of AI-based technologies. The proposal, a non-profit initiative created in 2020 in Finland by researcher Roberto Zicari, already has a total of 37 associated entities worldwide (26 laboratories, 9 institutions, and 2 research projects); among them CiTIUS, which in 2023 became the first Spanish institution to adhere to this new protocol.

 Z-Inspection® is presented as "a holistic system for assessing the reliability of intelligent systems," and has been developed in accordance with European Union standards, as a "checklist". The tool is based on the contributions of a multidisciplinary group of experts in different fields, whose task is to identify and discuss ethical issues and tensions in complex sociotechnical scenarios.

 CiTIUS researcher José M. Alonso, co-organizer of the 'Social Impact of AI' conference along with Professor Ramón López de Mántaras, emphasized the importance of the meeting "to integrate the work of the TELSEC4TAI network with Z-Inspection® and address the impact of AI in various fields, including education, economy, and health". Alonso, who in the fall of 2023 promoted CiTIUS' adherence to this new global initiative (the center became the first institution recognized by Z-Inspection® in Spain), stressed the significance of these conferences as "a milestone in promoting ethical and responsible AI, which will ultimately benefit society as a whole". Currently, there are 3 Spanish institutions adhered to Z-Inspection®: CiTIUS and UPC-Universitat Politècnica de Catalunya (both part of the TELSEC4TAI network); and the Ramón y Cajal Institute of Health Research (IRYCIS). All of them participated in the 'Social Impact of AI' conference.

 Reflection and Debate on AI

 The conference revolved around the debate on how AI can benefit citizens without creating a more divided and unsupportive society. In addition to representatives from all the research groups of the Spanish TELSEC4TAI network and the international initiative Z-Inspection®, the meeting also included members of the 'Interactive Natural Language for Explainable Artificial Intelligence' (NL4XAI) project, a pioneering project in training new doctors in the field.

 Led by CiTIUS and funded by the EU's Horizon2020 program, this network in 'Explainable AI' focuses on improving the explainability of intelligent systems; or, in other words, on how to improve machines' ability to explain "why" their decisions, in a language always understandable to humans.

 On the other hand, representatives of the Z-Inspection® initiative explained the details of the new process for evaluating the reliability of AI-based technologies, following the guidelines of the European Union High-Level Expert Group (HLEG) with the aim of achieving AI worthy of trust. This methodology has already been included in the OECD's Catalog of AI Tools and Metrics and is distributed under a Creative Commons license.

 An intense program to address a crucial issue

 The event began on May 30 with the opening by Miguel Jerez, advisor of the Ramón Areces Foundation. Next, CiTIUS researcher José María Alonso chaired the first session of the conference, which included presentations on the economic impact of AI, regulation of Artificial Intelligence, and misinformation, with speakers such as Luis Moreno Izquierdo, associate professor of Applied Economics at the University of Alicante (UA), Ricard Josep Martínez Martínez, director of the Microsoft Chair of Privacy and Digital Transformation at the University of Valencia (UV), and David Camacho Fernández, Professor at the Polytechnic University of Madrid (UPM).

 Magnus Westerlund, a professor at Arcada University of Applied Sciences in Finland, introduced Z-Inspection® and discussed a pilot on generative AI and higher education in the second session. In the afternoon, there was a meeting of the TELSEC4TAI project and a third session with presentations by Katarzyna Kaczmarek-Majer, a researcher at the Institute of Systems Research and Automation in Poland (IBSPAN), Javier Vázquez, a researcher at the Institute of Artificial Intelligence Research at the Polytechnic University of Catalonia (IDEAI-UPC), Miguel Ángel Sicilia Urbán, a researcher at the Ramón y Cajal Institute of Health Research (IRYCIS) in Spain, and José María Alonso (CiTIUS). The day ended with a round table chaired by Carles Sierra, director of the Institute of Artificial Intelligence Research of CSIC (IIIA-CSIC), in which Professors Senén Barro (CiTIUS), Luis Magdalena (UPM), and Alberto Fernández Gil (URJC) participated.

 Finally, the sessions on Friday, May 31, focused on the impact of AI in various sectors, including secondary education, higher education, and medicine, with speakers such as Zoe Falomir, associate professor at Umeå University (Sweden), Óscar Cordón, professor of Computer Science and Artificial Intelligence at the University of Granada (UGR), and María José Carreira, professor and researcher at the Singular Research Center in Intelligent Technologies of the USC (CiTIUS). The second session of the day addressed topics such as AI and language, ethics and philosophy of AI, and continuous monitoring of the social impact of AI, in which various approaches were presented through presentations by prestigious speakers such as Elena González-Blanco, Director of Artificial Intelligence for EMEA Digital Natives at Microsoft, Martín Pereira Fariña, Director of the Department of Philosophy and Anthropology at the University of Santiago de Compostela (USC), and Mariano Sánchez Martínez, Associate Professor at the University of Granada (UGR).