Artificial Intelligence and the Criminal Justice System: An Analysis of Liabilities and Consequences

Abstract

Artificial intelligence has become an influential technological force in contemporary criminal justice systems, reshaping criminal investigation, evidence analysis, predictive policing, sentencing, surveillance, cybercrime prevention, and institutional decision-making. This article analyzes the legal responsibilities and consequences arising from the use and misuse of artificial intelligence within the criminal justice system. The central problem addressed is whether traditional criminal law doctrines, which are historically based on human agency, intention, culpability, causation, and punishment, can adequately respond to autonomous or semi-autonomous AI systems that generate harmful, biased, or legally significant outcomes. Using a doctrinal, analytical, and interdisciplinary approach, the article examines the theoretical foundations of AI and criminal law, the liability of developers, programmers, users, operators, corporations, and public institutions, and the controversial question of whether AI itself can be treated as a subject of criminal liability. The analysis argues that direct criminal liability of AI remains conceptually weak because criminal punishment presupposes moral agency, consciousness, and blameworthiness. Instead, a layered model of responsibility is proposed, based on human and corporate accountability, foreseeability, control, negligence, compliance duties, and institutional oversight. The article also evaluates the broader consequences of AI for criminal justice, including algorithmic discrimination, opacity, privacy violations, mass surveillance, evidentiary uncertainty, due process risks, and the possible dehumanization of justice. At the same time, it recognizes AI’s potential benefits in improving investigative efficiency, detecting cybercrime, managing complex evidence, and supporting judicial administration. The article concludes that artificial intelligence should not be rejected from criminal justice, but its use must be strictly regulated through transparency, explainability, human oversight, anti-discrimination safeguards, auditability, and effective remedies. A human-centered legal framework is necessary to ensure that technological innovation strengthens rather than undermines accountability, fairness, and the moral legitimacy of criminal justice.