.png)
The AVAL Foundation
The Foundation was established with the aim of promoting and spreading the culture of validating artificial intelligence algorithms, particularly those used in decision-making processes that affect savings, citizens' rights and the public interest.
The Foundation works to obtain formal legal recognition of the need for independent algorithm validation procedures as an essential prerequisite for the responsible, transparent and reliable use of artificial intelligence technologies.
Purpose and activities
of social utility
The AVAL Foundation pursues social utility objectives and carries out its activities exclusively in the field of artificial intelligence, with particular reference to:
-
education, instruction and professional training, as well as cultural activities of social interest with educational purposes;
-
university and post-graduate education, including through collaborations with academic institutions and research institutes;
-
scientific research of particular social interest, aimed at developing methodologies, standards and criteria for validating algorithms.
Through these activities, the Foundation aims to contribute to the construction of an Artificial Intelligence ecosystem based on responsibility, verifiability and trust.
Paolo Savona
Founder, President
He is Professor Emeritus of Political Economy, Chairman of CONSOB (National Commission for Companies and the Stock Exchange) and former Minister for European Affairs in the Italian Government. He began his career at the
Research Department of the Bank of Italy, and undertook specialist studies at the Massachusetts Institute of Technology (MIT), under the guidance of Chrles P. Kindleberger and Franco Modigliani.
He was Minister of Industry, Trade and Crafts in the Ciampi Government, Secretary General for Economic Planning at the Ministry of the Budget, Head of the Department for European Policies of the Presidency of the Council of Ministers
and President of the Italian Guarantee System (twice).
He is a Knight Grand Cross of Merit of the Italian Republic, a Chinese Eisenhower Fellow in Taiwan and a Jemolo Fellow at Nuffield College, University of Oxford. He is also the author of numerous publications on monetary and financial issues and on economic dualisms. He has been a columnist for La Repubblica, Corriere della Sera, La Stampa, Il Sole 24 Ore, Il Messaggero, Milano Finanza, Formiche and Scenari economici.
He is Professor of Computer Science at the Luiss University, where he is also Director of the Master in Data Science and Management, the Master in Cybersecurity and the Master in Big Data and Management at the Luiss Business School. He has gained international experience as Visiting Professor at prestigious institutions such as Columbia University, the Université Paris-Sud, the Max-Planck-Institut fur Informatik in Saarbrücken and the Hong Kong University of Science & Technology, as well as being a Visiting Scientist at AT&T Research Labs and Microsoft Research. His research is predominantly focused on algorithms and in 2016 he was named a Fellow of the European Association for Theoretical Computer Science for his ‘fundamental contributions in the design and analysis of algorithms to solve theoretical and applied problems in graphs and large data sets, and for his role in establishing the field of algorithm engineering’. He has published over 300 scientific papers, advised major international companies and co-founded several technology start-ups. He has recently been appointed as member of the Italian Government commission on the application of algorithms to the editorial field, as the only technology expert. He is Dean of the newly established AI department at LUISS University.
Giuseppe Italiano
Founder, Vice President
Pro-Rector for Artificial Intelligence and Digital Skills, LUISS University
Michele Fratianni
Founder
He is Professor Emeritus at the Kelley School of Business of Indiana University (Bloomington, Indiana) and Professor of Economics at the Università Politecnica delle Marche (Ancona, Italy). He has held contemporary positions at the Catholic University of Louvain (1974-1991), Marquette University (1995, Distinguished Professor of International Economics), Free University of Berlin (1995, Bundesbank Professor of International Monetary Economics), as well as visiting positions at the Catholic University of Milan (1985) and the University of Rome La Sapienza (1991). He has also worked in government and international organisations, such as the U.S. Council of Economic Advisers (1981-1982, Senior Staff Economist) and the Commission of the European Communities (1976-1979, Economic Adviser), and has been a consultant to the Italian government (various ministries), the Bank of Italy, and the governments of Estonia, Latvia, Lithuania and Vietnam on domestic and international monetary issues. He is the Founder and current member of the Board of Advisors of Open Economies Review, Editor of the Rivista Italiana degli Economisti, and member of the Editorial Board of the Review of International Organisations. Professor Fratianni's research interests include international monetary economics, international trade, international economic policy and monetary history.
Kregel has been Professor of Finance and Development at Tallinn University of Technology in Estonia since 2006. He holds the position of Adjunct Professor at Johns Hopkins SAIS, where he co-directed the Bologna Centre in the late 1980s, and is a Visiting Professor at the University of Missouri-Kansas City. He is also a Senior Scholar at the Levy Economics Institute of Bard College. Until 2007, he was Head of the Branch of Policy Analysis and Development in the Office of Financing for Development of the United Nations Department of Economic and Social Affairs. Until 2004, he served as Senior Expert in International Finance and Macroeconomics at UNCTAD's New York Liaison Office, being its de facto Chief Economist. For many years, he held the Chair of Political Economy at the University of Bologna. Kregel studied mainly at Cambridge University (with Joan Robinson and Nicholas Kaldor) and Rutgers University (Ph.D. 1970 under Paul Davidson). He is a Life Fellow of the Royal Economic Society in London and in 2000 co-founded The Other Canon, a centre and network for research in heterodox economics, with principal founder and executive chairman Erik Reinert. Kregel is the programme director for the Master of Science in Economic Theory and Policy at the
Levy Economics Institute of Bard College, a programme launched in 2014.
Jan Kregel
Founder
.png)
OUR VALUES
SOCIAL UTILITY
Conscious use of algorithms, especially where they affect savings.
RIGOUR
Validation as a method, based on evidence, controls and quality.
INDEPENDENCE
Non-profit status and governance oriented towards the general interest.
COLLABORATION
Open dialogue with institutions, universities, businesses and communities.
FOUNDATION ACTIVITIES
DEVELOPMENT
Promotes the operational development of scientific research results through pilot projects aimed at implementing and integrating the frameworks and methodology studied.
SCIENTIFIC RESEARCH
Promotes and funds scientific research aimed at building the frameworks and assets necessary for the validation of algorithms. Promotes the training of those involved in the institutional use of artificial intelligence.
DISSEMINATION
Promotes the organisation of events aimed at exploring the topics under research, the publication of essays and scientific documents, and raising awareness among institutional actors of the need to regulate the phenomenon.
OUR VISION
In the 19th century, six serious financial crises in the United States, almost all of which were attributable to accounting fraud, highlighted the lack of reliable control and certification tools.
In response to this situation, Moody's launched the first financial statement certification activity in 1906, followed in 1945 by the creation of Standard & Poor's, marking the beginning of a structured system of validation and trust in financial markets.
Today, the spread of artificial intelligence is taking place in a similarly disorderly manner and without shared standards, comparable to that of nineteenth-century financial statements before
In the nineteenth century, six serious financial crises in the United States, almost all of which were attributable to accounting fraud, highlighted the lack of reliable control and certification tools.
In response to this situation, in 1906 Moody's launched the first financial statement certification activity, followed in 1945 by the creation of Standard & Poor's, marking the beginning of a structured system of validation and trust in financial markets.
Today, the spread of artificial intelligence is taking place in a similarly disorderly manner and without shared standards, comparable to that of nineteenth-century financial statements before their certification.
The current challenge of Artificial Intelligence
In Italy, important public institutions have declared their willingness to use artificial intelligence algorithms in decision-making processes. Among these, the Superior Council of Magistrates and the Revenue Agency have, however, raised a central question:
who is authorised to validate the algorithms before they are used?
At the international level, the Financial Conduct Authority has also recently announced the launch of a similar project, called Trust, aimed at defining new forms of reliability and control of algorithms, although the operational details have not yet been disclosed.
The role of the AVAL
Foundation
In this scenario, the AVAL Foundation acts as a cultural, scientific and institutional reference point for defining principles, methods and validation processes for artificial intelligence algorithms, helping to fill an increasingly evident regulatory and operational gap.
The validation of algorithms is not only a technical issue, but also an essential safeguard of legality, fairness and protection of the collective interest.
Algorithm validation as the foundation of reliability
VALIDATION IS A MUST.
BUT WHY IS IT SO DIFFICULT TO ACHIEVE?
Algorithms face critical challenges such as a lack of transparency, which hinders error detection and accountability, biases in training data that can lead to unethical outcomes, and poor generalisation to real-world scenarios, which limits their reliability.
WHY DOES IT REQUIRE A METHOD THAT INTEGRATES TECHNICAL RIGOUR, TRANSPARENCY, ETHICAL GUARANTEES AND ADAPTABILITY?
‘Validity’ is not a given: it must be defined and scientifically argued through metrics that accurately represent the intended use and associated risks. This requires adequate and controlled reference data, reproducible experimental protocols, and the quantification of uncertainty and margins of error. Metrics must demonstrate generalisability and stability with respect to changes in population, context and time, including monitoring and updating.
THE FOUNDATION FOR THE VALIDATION OF
ARTIFICIAL INTELLIGENCE ALGORITHMS
It intends to implement this response by giving a significant boost to scientific research, the dissemination of the culture of validation, and its regulation at the legislative and operational levels.
HOW TO SUPPORT AND PARTICIPATE IN THE FOUNDATION'S ACTIVITIES
FINANCIAL DONATIONS AND SPONSORSHIPS
Direct financial support for the Foundation's priority initiatives and programmes and visibility partnerships for events, campaigns and projects with scientific impact.
JOINT RESEARCH AND TECHNICAL SUPPORT
Collaborations on shared research projects and
contributions in terms of know-how, digital tools and specialist assistance for operations.
CONTRIBUTION TO OUTREACH ACTIVITIES
Preparation of scientific documents and publication of articles for the specialised press.
CONTACT
For any general inquiries, please fill in the following contact form:
BACK TO TOP