Loading...

Prof. Stavroula Tsinorema

Prof. Stavroula Tsinorema

Prof.Stavroula Tsinorema

Prof. Stavroula Tsinorema
University of Crete, Greece


Title: Building an Ethical AI. Which Principles, Whose Values?

Abstract: The demand for ethical reflection and robust policies in AI governance has given rise to a plethora of world-wide initiatives to establish ethical principles, rules, recommendations, codes, to meet practical challenges and prospective risks. However, the huge variety of proposed sets of principles frequently tends towards repetition, overlap, lack of coherence, ambiguity. We need a robust, harmonised ethics framework for the governance of AI, implemented by trustworthy institutions. But how is it to be crafted? Which principles and whose values should guide action? Firstly, robust conceptual analysis needs to be undertaken. Principles and values differ in their logical properties and play a different role in practical reasoning. Principle-based reasoning and value-based reasoning differ in their modalities and satisfy different coherence criteria. Both, however, are indispensable in AI ethics. A comprehensive framework of principles is required to constrain the AI race globally. In its concretization, it needs to be enabled by value-based modes of reasoning in diverse contexts and domains of application. A frame of fundamental principles of cosmopolitan scope (including human rights), in alignment with value pluralism (not relativism), can furnish a solid basis for AI governance and oversight. In addition, as neither principles nor values can provide instructions for their application, the latter must always be supplemented by human judgement. An operational or functional morality, built into the AI systems themselves, will not suffice. Reasoning ethically is always a task rather than an automatic process. This has crucial implications regarding the “control problem”, which will be highlighted and put under normative scrutiny. The nature of ethical thinking imposes constraints on the capability and acceptability of AI systems “making ethical decisions”.

BIO: Stavroula Tsinorema is Professor of Philosophy and Bioethics, Founding Director of the Interinstitutional Postgraduate Programme “Bioethics” and Director of the Centre for Bioethics of the University of Crete. She is member the Hellenic National Commission for Bioethics and Technoethics and member of the Hellenic National Committee for the Welfare of Animals Used for Scientific Purposes. She holds a First Class Honours Degree in Philosophy from the University of Athens, followed by M.A. in Moral and Social Philosophy and Ph.D. in Moral Philosophy from the University of Exeter. Her research publications focus on the epistemology of value-concepts, moral and social philosophy, bioethics, ethics of science and technology, research ethics, environmental ethics, Kant’s philosophy, Ludwig Wittgenstein. She sits on the Institutional Ethics Committee of the European Union Reference Centre for the Welfare of Aquatic Animals-EURCAW Aqua and she is the National Representative for the ERA Action on Research Ethics and Integrity of the European Commission. She chairs the Research Ethics Committee of Hellenic Pasteur Institute. She has served on a number of national and European Committees, including as member of the European Commission’s “Expert Group to advise on specific ethical issues raised by driverless mobility”, which issued the «Independent Expert Report on the Ethics of Connected and Automated Vehicles, 2020»; founding chair of the Research Ethics Committee of the University of Crete; member of the National Committee for Deontology for Clinical Trials of the Ministry of Health. She has also served as Dean of the Faculty of Philosophy of the University of Crete.