Agentic AI, Retrieval-Augmented Generation, and the Institutional Turn: Legal Architectures and Financial Governance in the Age of Distributional AGI

This research paper examines how agentic AI systems that can act autonomously challenge existing legal and financial regulatory frameworks. The authors argue that AI governance must shift from model-level alignment to institutional governance structures that create compliant behavior through mechanism design and runtime constraints.

• →Agentic AI systems with autonomous capabilities present unprecedented challenges to current legal and financial regulatory frameworks.
• →Traditional AI alignment through training-time interventions like RLHF is insufficient for persistent AI agents.
• →AI governance must evolve from isolated model behavior to institutional environments with runtime governance graphs.
• →Retrieval-Augmented Generation combined with agentic systems creates new implications for legal accountability.
• →Compliant AI behavior should emerge through carefully designed payoff landscapes rather than internalized values.