A Comparative Analysis of Legal Liability and Accountability Frameworks for Autonomous Artificial Intelligence

Artificial Intelligence (AI) has evolved from a supportive tool into a semi-autonomous system
influencing sectors such as healthcare, finance, transportation, and digital communication. This
transformation raises significant legal concerns, particularly regarding liability when AI systems
cause harm. Traditional legal frameworks are structured around identifiable human or corporate
actors, making it difficult to assign responsibility in cases where AI operates with a degree of
autonomy and unpredictability.3
This study undertakes a comparative analysis of legal liability and accountability frameworks for
autonomous AI across major jurisdictions, including the European Union, United States, United
Kingdom, China, and India. It argues that neither fault-based liability nor the concept of granting
legal personhood to AI offers a complete solution. Fault-based approaches face challenges due to
the opacity of AI systems, while AI personhood lacks practical feasibility in ensuring compensation
and accountability.
The paper proposes a layered regulatory approach that combines ex ante obligations—such as risk
assessment, documentation, and monitoring—with ex post liability mechanisms tailored to the level of risk involved. It also emphasizes the need for evidentiary flexibility and compensation mechanisms, including insurance frameworks, especially for high-risk AI applications. The study concludes that India must adopt a proactive and structured regulatory model that ensures accountability across the AI lifecycle while safeguarding innovation. Such a framework would bridge the accountability gap without conferring independent legal status on AI systems.4