UC3M and Universia obtain an ENIA Chair in Artificial Intelligence in Data Economy

- ES- EN

The Universidad Carlos III de Madrid (UC3M) is one of 22 institutions that have been selected by the Ministry for Digital Transformation and the Civil Service to create an ENIA Chair to further the development of artificial intelligence (AI)-based applications. The AImpulsa UC3M-Universia Chair, as it is called, will be the only one of its kind in Spain in the area of Data Economy and will collaborate with Universia-Banco Santander, through Santander Universities.

The ENIA Chairs’ objectives, which depend on the Secretary of State for Digitalisation and Artificial Intelligence of the Ministry for Digital Transformation and Public Administration, include promoting research and knowledge transfer in AI in different areas of knowledge and encouraging the promotion of the professional and university offer aimed at the development and innovation of this technology. They also focus on public-private collaboration and its sustainability.

"It is important to create synergies between the university and the company in order to generate maximum value by transforming theoretical research into tangible innovation within the company, and Universia will provide an advanced technological environment, real use cases and a team of experts, both in the corporate world and in AI," says José Manuel de la Chica, director of Generative AI at Banco Santander.

The resolution of the Ministry provides for the admission of 15 national and 7 international chairs, which will address 10 areas of knowledge: aeronautics and aerospace, agriculture, green algorithms, health sciences, sustainable development, data economy, responsible and ethical AI, music and arts, demographic challenge and language technologies.

The AImpulsa UC3M-Universia Chair, which will include a multidisciplinary representation of researchers from the University, experts in the field from the financial institution and other international scientists, aims to address the challenges of the exploitation of personal data by large technology companies and the ethical implications of data privacy. One of the objectives of its research programme is to create a new personal data economy that is transparent, fair, inclusive and responsible.

"This will make it possible to maintain and improve current economic incentives, while reducing the harmful impact on the most vulnerable people and communities, promoting a new and healthier productive economic fabric based on this new data economy that we intend to explore during the course of the chair. On the other hand, at IBiDat, we are immersed in a disruptive line of research, focused on developing models and algorithms linked to AI that are interpretable and unbiased. The results will be tested in the very stimulating collaboration we have with Universia", says the head of the AImpulsa Chair, Rosa Elvira Lillo, a professor of Statistics and Operations Research at UC3M and director of the big data research institute IBidat (uc3m-Santander Big Data Institute).

According to the creators of this Chair, in the era of the digital economy in which we find ourselves, large technology companies are accumulating huge amounts of personal data, driving big profits and generating a new data economy, according to the creators of this Chair. An OECD study shows that in the United States alone, these data-based companies had a turnover of more than $60 billion in 2017. In Canada, the amount was around 1.4 billion dollars and in the European Union it was estimated at between 19 and 50 billion euros in 2016. "In parallel to this phenomenon, deep and growing concerns are emerging about privacy, ethics and fairness in the use of this data and in the algorithms that use it to learn and predict," explains another of the researchers participating in the AImpulsa UC3M-Universia Chair, Rubén Cuevas, Associate Professor in UC3M’s Telematics Engineering Department and deputy director of IBiDat.

That is why the AImpulsa UC3M-Universia Chair on Data Economy and Responsible Applied Artificial Intelligence for the Creation of Exponential Value proposes an initiative with an ambitious and advanced comprehensive and technical focus, according to the researchers. Specifically, they have the ambition to be at the forefront of innovation in the application of solutions to ensure interpretability and fairness in AI algorithms by developing advanced and large-scale models. To do this, they will use deep neural networks, reinforcement learning algorithms and natural language processing, as well as simulation models, mechanism design and experimentation of large social and economic systems.

Their research programme also includes a novel economic analysis of the impact of applying responsible algorithms, with the aim of encouraging the adoption of this type of solution by big data and AI technology companies. Instead of using traditional approaches such as strict regulations or sanctions (whose results may be limited or even counterproductive in some complex scenarios being generated by the new data economy), the researchers propose to develop innovative and scalable economic incentives that encourage companies to adopt and promote responsible AI in the medium and long term. To do this, they will design economic models and mechanisms that take advantage of the most advanced techniques in game theory, mechanism design, economics and information theory and computer privacy to create systems that encourage the adoption of responsible AI, and also penalise the use of algorithms that are harmful to society.

"All of these challenges, and all those appearing on the horizon in relation to the new AI economy, will be continuously explored within the Chair, demonstrating once again that knowledge transfer is fundamental in industry, something we at Universia are firmly convinced of," says De la Chica.

"At UC3M’s IBiDat, we are convinced that the AImpulsa chair will provide significant developments in aspects of AI that have not been explored until now, which will provide more confidence throughout the debate around the use of these methodologies, since our purpose is to move towards a fairer, more explainable and more efficient AI," concludes Lillo.