LEGAL AND ETHICAL LIMITS OF ARTIFICIAL INTELLIGENCE IN THE ECONOMY: CHALLENGES OF REGULATORY POLICY

Abstract

The article analyzes the legal and ethical problems of introducing artificial intelligence (AI) technologies into the economy, which arise due to the fact that technologies are developing faster than the formation of a relevant regulatory framework. Based on the analysis of Russian legislation (the framework Federal Law «On Supporting the Development of Artificial Intelligence Technologies in the Russian Federation» adopted in 2026, the Code of Ethics of the Bank of Russia), international regulatory approaches (the European AI Act, the strategies of the US and China) and scientific research, key ethical challenges are identified: algorithmic bias and discrimination, the «black box» problem and the lack of explainability of algorithms, the blurring of responsibility for the decisions of AI systems, threats to privacy and behavioral manipulation. It is substantiated that soft regulation through ethical codes creates the risk of «ethical Washington» (declarativeness), while excessively strict standards can hinder innovation. Suggestions for improving regulatory policy are formulated: the introduction of a risk-oriented approach to the classification of AI systems; Legislative codification of requirements for the explainability of algorithms for high-risk areas (credit scoring, insurance, public administration); the creation of an institution for independent audit of AI systems; the development of mechanisms for the liability of developers and owners; strengthening the protection of personal data during neural network training; the development of mechanisms for public oversight and international coordination. A conclusion is drawn regarding the need for a transition from reactive to proactive regulation capable of anticipating systemic risks, ensuring a balance between stimulating technological development and protecting citizens' rights.