Artificial Intelligence Liability and Legal Personality in the Age of Autonomous Systems

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Artificial intelligence (AI) is rapidly transforming modern society, with increasingly autonomous systems making decisions that affect healthcare, finance, transport, commerce, criminal justice, and public administration. As AI systems evolve from merely executing programmed instructions to learning, adapting, and acting with limited human intervention, traditional legal doctrines of liability are under growing strain. Existing legal frameworks generally attribute responsibility to human actors, including programmers, manufacturers, operators, and users. However, these approaches become increasingly inadequate when harm results from autonomous decision-making that cannot easily be traced to a single human actor. This article critically examines whether AI should be recognised as possessing a limited form of legal personality capable of giving rise to legal liability. Drawing on doctrinal legal research and comparative and interdisciplinary analysis, the study evaluates existing theories of legal personality, corporate personhood, agency, product liability, and negligence, as well as emerging international regulatory approaches, including the European Union AI Act. It analyses whether the concept of legal personality should evolve to accommodate highly autonomous AI systems while preserving the fundamental objectives of accountability, justice, and effective compensation for victims. The article argues that granting AI unrestricted legal personality would undermine established principles of legal responsibility and obscure human accountability. Nevertheless, the law should recognise a functional and limited legal status for specified categories of highly autonomous AI operating within tightly regulated environments. The article concludes that the future of AI regulation depends not on treating machines as equivalent to natural persons but on developing an adaptive legal framework that balances innovation, accountability, public trust, and the rule of law in an increasingly autonomous digital society.