Interdisciplinary Remarks on the Assessment List on Trustworthy AI (ALTAI) Applied to a P5 Medicine Tool

Abstract

The authors apply the Assessment List on Trustworthy AI (ALTAI) to a possible AI-based tool aiming at supporting the melanoma cancer diagnosis. They take the opportunity to provide an interdisciplinary analysis of the proposed selfassessment tool in light of its possible mandatory application in R&D&I. The presented empirical exercise highlights some pros and cons of the adopted checklist, stimulating further remarks on the EU regulatory initiatives on AI. Finally, we try to understand the improvement of procedures, medical knowledge and treatment collected and improved during these months, that allowed for a lower mortality rate in the referring period.