Quale diritto penale dell’IA? Alcune riflessioni a partire dalla proposta di regolamento dell’Unione Europea
digital
![]() Articolo
€ 6,00
|
Ebook in formato Pdf leggibile su questi device:
|
|
The paper investigates the possible shape of the criminal law of Artificial Intelligence. By relying on an approach which does not recognize a direct form of criminal responsibility of the AI systems, the work examines the proposed Artificial Intelligence Act by the European Union – which, to date, has not been finally approved yet – and the risk-based approach adopted. Firstly, the prohibited AI practices are examined, with particular attention to the ‘real-time’ remote identification systems, whose discipline has been changed during the legislative process. Then, the paper focuses on high-risk AI practices, which require different risk management obligations on those who use them, depending on their role. Based on such an analysis, the work stresses that a risk-based approach, which relies on different rules to reduce the risks associated with AI systems, can serve as a starting point to acknowledge the criminal liability of individuals and corporations.
keywordsAI Act; AI criminal law; risk-based approach; Bruxelles effect; biometric identificationBiografia dell'autoreUniversità degli Studi della Campania “Luigi Vanvitelli”. Email: marco.colacurci@unicampania.it. |
Consulta l'archivio
Articoli Open Access
Ultimi 3 numeri
Annate disponibili online
2025202420232022202120202019201820172016201520142013201220112010200920082007200620052004200320022001200019991998199719961995199419931992199119901989198819871986198519841983198219811980197919781977197619751974196919681967196619651964196319621961196019591958195719561955195419531952195119501943194219411940