Autonomous AI Agents in Regulated Financial Services

Governance Framework for Agentic Systems Under DORA and the EU AI Act


The first published governance framework for autonomous AI agents in regulated financial services. Three tiers of agent autonomy. Article-by-article DORA analysis identifying thirteen provisions where the regulation fails for agentic systems. Resolves the structural paradox between AI Act Article 14 human oversight and autonomous operation.

Six-category Agentic Risk Taxonomy (ART-1 through ART-6). Three uniquely agentic with no analogue in compositional risk. Four architecture-specific compliance profiles for trade execution, AML investigation, claims processing, and customer onboarding.

Files

File Description
paper.pdf Full governance framework

Citation

DOI: 10.6084/m9.figshare.31516171

License

CC BY-NC-ND 4.0