PREDICTIVE JUSTICE: AN INTERDISCIPLINARY APPROACH BETWEEN PHILOSOPHY OF LAW, LEGAL COMPARISON AND INFORMATICS
CONVENORS: STEFANO GUERRA, SIRIO ZOLEA
Currently, the ability to forecast the outcome of a judgment relating to a specific case through algorithms based on jurisprudential, normative and factual data is called predictive justice (part of the digital justice – AI applied to justice – which includes in its “hard” meaning the automation of algorithmic decisions in a judgment). Over the last few years, with the advent of Big data and the development of machine learning systems, predictive justice has returned to the center of a multilevel worldwide debate involving scholars of various disciplines (philosophy, law, informatics) and practitioners of various sectors (enterprises, courts and public administrations). Therefore, in-depth studies are needed about the impact that computational heuristics can have on the complexity of law. The concrete application of predictive justice implies both epistemic issues and potential dangers for fundamental rights.
In this regard, the contribution of legal comparison is paramount to understand and critically assess the experiences of various national systems that already adopt or are trying out different models of predictive justice. Several examples can be found on a global scale: China (four online databases of judicial information established by the Supreme People’s Court), some North American States (Correctional Offender Management Profiling for Alternative Sanctions – COMPAS), Estonia (e-Estonia), Netherlands (Rechtwijzer System), UK (Harm Assessment Risk Tool – HART), France (Predictice and other LegalTech companies), Italy (experimental collaborations between some jurisdictions and universities), which have been severely criticized by scholars, who emphasize the critical points and the weaknesses of algorithm fairness. Also, the national legislators are starting to become aware of the issues of predictive justice, such as in France, from the one side engaging public institutions to grant everyone interested online open access to all judgments, from the other side forbidding and sanctioning the more hazardous employments of algorithms, for example those aiming at individually profiling judges. In the EU legal debate, the Proposal for a Regulation laying down harmonized rules on AI (AI Act), establishing legal and ethical limits, could be very relevant in the next future, while the GDPR, concerning data protection, is already operating to limit potential abuses caused by predictive justice.
In short, predictive justice can produce benefits, potentialities and risks, raising serious legal, ethical and informatics concerns. Hence, this workshop aims at an in-depth multi, inter and trans-disciplinary analysis on predictive justice, providing a platform to discuss on the evolution of its technical applications and experiments, on its theoretical interpretations and implications in philosophy, law and informatics, and on the opportunity of integration between predictive justice and human (interpretative) justice.
The organization of the workshop will involve scholars of various disciplines (including, in particular, legal philosophy, comparative law and informatics), working at the crossroads of artificial intelligence, law and justice, both by direct invitation and by an open call for papers. Participants will have the opportunity to present recent developments in these fields and to discuss the main challenges lying ahead, both theoretically and practically, promoting a cross-cutting dialogue.
Stefano Guerra – Ph.D. in Philosophy of Law
(SIFD – Italian Section of IVR; University of Macerata)
Sirio Zolea – Ph.D. in Comparative Law
(Faculté Libre de Droit, d’Économie et de Gestion de Paris ; University of Teramo; University of Macerata)