Fairness (in particular procedural fairness) is not only an abstract concept reserved for academics and regulators. It is a concept with real-life implications that people from all venues of life reflect on, especially in the face of new technologies such as AI.
For the purpose of starting a general conversation on fairness from a variety of vantage points and for the purpose of sharing insights and information beyond the confines of academia, the JUST-AI ‘Meet the Expert’ Podcast offers a unique access to audiences from around the world to discussions with experts from various fields on the ways in which AI affects and is likely to affect our understandings of fundamental societal concepts like intelligence, humanity, democracy, fundamental rights and of course, fairness.
How can robots and AI systems enhance the state of our health and promote inclusivity? How can care robots and recruitment algorithms be designed in order to truly enhance human abilities and equality? These are in essence the issues that Eduard Fosch Villaronga seeks to answer through the groundbreaking research he conducts within the Safe & Sound and BIAS projects.
No one is better placed to talk about the tumultuous negotiations surrounding the AI Act than Luca Bertuzzi, an independent journalist in the field of new technologies. In this episode of the JUST-AI Jean Monnet Podcast, Luca gives us a glimpse behind the scenes, astutely commenting on the ratio legis of the AI Act, but also highlighting possible shortcomings in this Regulation’s future application.
As Artificial Intelligence (AI) takes on a greater role in decision-making, whether fully-automated or not, more and more voices have been raised to highlight just how discriminatory these systems can be. Foremost among these is Raphaële Xenidis, Assistant Professor at Science Po, Paris. As a true expert in discrimination law, she uses her analytical framework developed to make a precise a precise diagnosis of the risk associated with algorithmic biases. So sit back and relax, as Raphaële explains why AI systems are inherently bias, and why debiasing them isn't as easy as it might first seem.
If you are a migrant, worker, civil servant, or hold any role where AI systems process your data such as country of origin, gender, or age, does the law provide procedural mechanisms allowing you to verify the accuracy of that processing? The critical themes of AI accuracy and trustworthiness, effective judicial protection, and the dignified treatment of individuals affected by AI-driven decisions were at the heart of this episode of the JUST-AI Jean Monnet Podcast featuring brilliant Francesca Palmiotto, Professor at IE University, Madrid.
With a rich background in criminal law and procedure, Prof. Chiao’s original and insightful scholarship sheds light on the critical role of discretion—both for regulators when enacting procedural frameworks, and, perhaps more importantly, for judges when incorporating AI systems into criminal sentencing. He predicts that as AI systems become more robust, they could assist courts in delivering fairer verdicts. However, this raises an open question: how should we reinterpret and apply due process rights for litigants when courts integrate AI into their decision-making?
When Bentham first conceptualized the panopticon - a metaphor for an all-seeing disciplinarian - he could not have foreseen the profound impact data would have on the nature of surveillance centuries later.
In today’s episode, Prof. Maša Galič explores how, in the context of smart cities and living labs, datafication has transformed mass surveillance: shifting its purpose from discipline to care, and involving the active, willing participation of those being observed. This shift, however, raises interesting legal questions.
In an era of pervasive surveillance and data sharing, should the notion of personal privacy give way to group privacy?
In criminal procedure, can AI tools used to collect data in investigations be trusted as reliable instruments for fair and informed court decisions?
These are only some of the many issues Prof. Galič addresses in today’s discussion. So join us for this insightful and thought-provoking conversation, as Prof. Galič talks about her work and challenges us to rethink surveillance, privacy, and justice in the digital age.
What level of privacy protection can be reasonably expected when personal data is processed? This is the central question Dr. Paul Friedl explores in his latest book, Reasonable Expectations of Privacy: With Special Regard to European Privacy and Data Protection Law.
The issue is two-fold within normative frameworks like the GDPR, the EU Charter of Fundamental Rights and the ECHR. On one hand, individuals may reasonably expect that the processing of their personal data - even when based on consent - respects their right to privacy. On the other hand, the scope of that processing can sometimes exceed what one might have reasonably expected.
Drawing the line between lawful data processing and the expectation of privacy it warrants has been a longstanding challenge for both the European Court of Human Rights (ECtHR) and the Court of Justice of the European Union (CJEU). These courts have worked to clarify not only what individuals actually expect, but also what they are legally entitled to expect under privacy-protecting frameworks, in instances where their data is being processed.
Dr. Friedl’s book offers a careful and insightful analysis of this evolving case law and is a must-read for anyone interested in data protection, privacy, and emerging technologies.
Join us as he shares his findings on this episode of the JUST-AI JMCE Podcast.
