Marcel Osmond

Marcel Osmond, Thomas Jego

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.

Marcel Osmond

The proliferation of agentic artificial intelligence systems--characterized by autonomous goal-seeking, tool use, and multi-agent coordination--presents unprecedented challenges to existing legal and financial regulatory frameworks. While traditional AI governance has focused on model-level alignment through training-time interventions such as Reinforcement Learning from Human Feedback (RLHF), the deployment of large language models (LLMs) as persistent agents necessitates a paradigm shift toward institutional governance structures. This paper examines the intersection of agentic AI, Retrieval-Augmented Generation (RAG), and their implications for legal accountability and financial market integrity. Through analysis of the Institutional AI framework, we argue that alignment must be reconceptualized as a mechanism design problem involving runtime governance graphs, sanction functions, and observable behavioral constraints rather than internalized constitutional values[...].The analysis concludes that the future of AI governance lies not in perfecting isolated model behavior, but in architecting institutional environments where compliant behavior emerges as the dominant strategy through carefully calibrated payoff landscapes.