Ethical and Legal Risks of Using Artificial Intelligence in Law Enforcement and Special Services

Abstract

The article examines the main practices of using artificial intelligence (AI) in law enforcement and special services and identifies key ethical and legal risks arising from the deployment of predictive models, big data (Big Data) analytics, and biometric technologies. It is shown that AI can enhance the operational capacity of prevention and investigation by integrating fragmented information resources, rapidly processing large datasets, supporting operational decision-making, and partially automating routine pre-trial investigative procedures. At the same time, it is established that working with Big Data may reproduce and amplify biases embedded in the data and in data-collection practices, producing discriminatory effects and algorithmic profiling of groups, as well as creating risks of disproportionate interference with privacy. These threats are particularly acute in practices of forecasting high-risk areas, prioritizing resources, and deploying facial recognition in public spaces, where technological “accuracy” alone does not ensure legality or fairness. The article emphasizes that statistical patterns and predictive assessments cannot replace legal evaluation, evidentiary standards, or individualized decision-making; AI outputs must remain auxiliary, while responsibility for decisions and interferences with human rights should rest with an authorized law-enforcement actor, provided that adequate procedural safeguards and avenues for appeal are available. The article concludes that permissible use of AI in law enforcement and special services is possible only under clearly defined limits of deployment, effective oversight and redress mechanisms, and the preservation of the primacy of human dignity and the rule of law.