Artificial General Intelligence: a New Chapter in the Global Legal Discourse

Abstract

This article offers a comprehensive analysis of the legal and socio‑civilisational implications associated with the potential emergence of Artificial General Intelligence (AGI) a system capable of performing general cognitive functions at or beyond the level of human intelligence across an unrestricted range of domains. It argues that the transition from narrow AI to AGI represents not merely a technological milestone but a profound civilisational shift, one that fundamentally challenges the adequacy of existing legal frameworks designed for earlier generations of AI systems. The article examines the technological prerequisites for AGI development, including large‑scale model architectures, multimodality, and autonomous agentic systems, and systematises contemporary forecasts regarding the likely timelines of AGI’s arrival. The potential socio‑economic consequences of AGI are analysed across three key domains: scientific progress, labour market transformation, and public administration. Particular attention is devoted to risks for fundamental human rights, including the right to privacy, the principle of non‑discrimination, freedom of expression, and the integrity of democratic processes. A separate section traces the evolution of corporate AI safety standards from the existential‑risk framing of 2015 to the narrowly legalistic liability‑management approaches of 2025 demonstrating the structural insufficiency of industry self‑regulation as a mechanism for governing risks of AGI magnitude. The article critically evaluates the principal international legal instruments in this field, including the EU AI Act, the Council of Europe Framework Convention on AI, the OECD AI Principles, and the UNESCO Recommendation on the Ethics of AI, and considers the implications of the European Commission’s withdrawal of the proposed AI Liability Directive. The article concludes that a qualitative regulatory shift is required: from fragmented governance of individual narrow‑AI applications toward a systemic legal regime for frontier‑level technologies. Such a regime must include mandatory human‑rights impact assessments, independent technical audits, incident‑reporting obligations, and authorisation procedures for models presenting systemic risk. A hybrid model of legal liability for AGI developers and operators is proposed, while the attribution of legal personality to AGI is characterised as premature and potentially harmful. Instead, the article advances the concept of AGI as an object of heightened legal scrutiny.

Ähnliche Arbeiten