AI systems are reshaping society but challenge traditional criminal law, which relies on human intent and action. Globally, the EU uses risk-based rules; the U.S. applies existing laws with sentencing enhancements; the UK and Singapore adapt negligence and governance frameworks. Liability models range from holding humans responsible to strict liability for high-risk AI. India currently applies old criminal and IT laws, with courts emphasizing rights and oversight. Scholars propose negligence rules, strict liability, and mandatory insurance for AI harms. Policymakers are considering risk-based laws, sandboxes, standards, and human oversight to build a clear, future-ready AI liability framework.
