This article outlines and analyses critically the challenges that artificial intelligence (AI) poses to criminal law and how existing Greek law deals with them. It gives a definition of AI under Greek criminal law and addresses the suitability of existing criminal laws to deal with AI-related offences, including the challenges posed to the requirement of actus reus, mens rea and causality. Engaging with the idea of introducing e-legal personhood to deal with AI, it argues that criminal law rests inextricably upon human agency, and instead suggests a model based on the notion of objective negligence for AI criminal liability. Finally, it explores briefly complementary non-criminal measures and alternatives to criminalisation when regulating abuse or harmful use of AI.