This paper proposes a doctrinal framework for attributing criminal responsibility in human-AI systems, focusing on human-machine collaboration through the concept of hybrid intention. Hybrid intention exists when a human initiates a purposive course of conduct, the realisation of that purpose is constitutively mediated by an artificial system that contributes more than a mere instrument, and a limited yet assessable veto or control window is available to the human agent. A three-part attribution test follows from this structure and aligns with continental principles of legality and guilt while avoiding the personification of machines. The test directs courts to examine initiation, technological mediation and the existence and use of a control window, drawing on evidence of foreseeability, traceability and realistic opportunities to intervene. The paper develops the test through a hypothetical case study of surgical AI in which a hospital deploys an AI-assisted system for laparoscopic procedures and a manoeuvre based on real-time imaging results in a vascular injury. The framework provides court-usable criteria for present technologies and informs compliance by design for regulators and developers while preserving anthropocentric culpability and narrowing responsibility gaps in posthuman contexts.