GOVERNED AUTONOMY
Research on computational governance in regulated financial services
Governed Autonomy examines how financial institutions can govern AI in practice: what a system may do, who authorises it and what evidence supports each decision. The research combines regulatory analysis, reference architectures and a free diagnostic, with an initial focus on Singapore private banking.
Published research
Governed Autonomy
How Private Banks Can Industrialise AI Within Regulatory Boundaries
Published 25 March 2026 · Strategic architecture paper
A proposed architecture for governing AI in Singapore private banking. The paper connects regulatory analysis, bounded agent authority and human accountability through a periodic client review—from gathering evidence to approving communication and retaining the decision record.
Independent analysis of public sources. The architecture is a research proposal; economic claims are directional.
Read the paper and publication detailsResearch tools
AI Governance Readiness Assessment
Available · Free research diagnostic
A self-assessment for Singapore private banks, using 193 assessment criteria across seven domains drawn from 13 source instruments. The corpus distinguishes supervisory expectations, consultation proposals, observed practices, methodologies and assurance standards. Results identify areas for review; they do not establish legal compliance or constitute certification.
Explore the free diagnostic and its methodologyReference architecture
Working draft · Illustrative example
The proposed design separates operational data, institutional permissions and regulatory requirements. A fictional client review illustrates how evidence, approval and an accountable decision record fit together. This is an inspectable design, with implementation and evaluation still to be completed.
Read the architecture and worked exampleResearch in progress
Governed Autonomy II
In progress
Examines accountability when autonomous agents transact through tokenised financial infrastructure, using an autonomous robotaxi as the exemplar. The question moves from governing AI within an institution to governing the agent and the infrastructure together.
Governed Autonomy III
Planned
Investigates regulatory system design: how supervisory boundaries, jurisdiction and institutional accountability should respond when rules and their implementation become computational.
About the programme
Governed Autonomy is led by Adrien Pesa, a Singapore-based governance practitioner and researcher. He previously served as Global Head of Compliance at Arta Finance and Chief Compliance Officer of Stash Investments. His research draws on experience designing and operating governance frameworks for digital wealth platforms across the United States and Singapore.
He is the author of Governed Autonomy and the developer of the programme’s free research diagnostic.
The programme is self-funded and non-commercial—no entity, no clients, no revenue.
The programme is open to research collaboration.
Research enquiries: contact@governed-autonomy.com