This webinar explores the degree to which AI "entities" can themselves be held liable for criminal activity.  It will include an overview of the fundamental processes by which contemporary Large Language Models (LLMs) operate, framing these capabilities in the context of how the criminal law views volitional action. The session uses the question of AI criminal liability to illustrate for a legal audience the capabilities and limitations of current AI models, how they impersonate human-like activity, and how the application of that understanding can inform questions across other areas of legal practice.