🤖 AI Summary
This study addresses critical misalignments between current legal frameworks and the technical realities of AI agents. Existing regulations often conflate model capabilities with agent architectures, misinterpret probabilistic predictions as cognitive decisions, and misunderstand the nature of autonomy, thereby failing to effectively govern the actual operational mechanisms of AI systems. Through a systematic comparison of eleven regulatory texts—including the EU AI Act, OECD/G7 principles, and NIST guidelines—against six prevalent AI agent development architectures, this work employs literature analysis, cross-regulatory comparison, and architectural dissection to uncover structural discrepancies between legal definitions and technical implementations. It proposes, for the first time, a consensus-based technical definition grounded in developer documentation that explicitly incorporates key autonomy-enabling elements such as system prompts, API permissions, sandboxing mechanisms, and orchestration code, offering both a theoretical foundation and a practical framework for precise and effective AI regulation.
📝 Abstract
This article presents the first systematic comparative survey of how public bodies, international organisations, national regulators, and the private sector define agentic artificial intelligence, identifying the technical inaccuracies pervading each definition. Analysing eleven regulatory instruments and industry frameworks -- including the EU AI Act, the OECD/G7 Principles, NIST, the UK ICO, and the European Commission -- alongside six leading developer architectures, this study demonstrates a persistent definitional gap: legal definitions consistently conflate model capability with agentic architecture, attribute cognitive deliberation to probabilistic token prediction, and treat autonomy as a scalar property rather than a structural shift from single-inference to iterative execution loops with tool integration. A consensus technical definition synthesised from developer documentation is proposed. The article examines the consequences of this gap, demonstrating that definitional imprecision produces regulatory instruments structurally incapable of governing the actual mechanisms -- system prompts, API permissions, sandboxing, and orchestration code -- that constitute agentic autonomy.