AI governance does not fail at approval. It fails when drift, workload, and accountability pressures appear after deployment. This paper outlines a practical integrity architecture for governing AI in real clinical systems.
Dr Alwin Tan, MBBS, FRACS, EMBA (University of Melbourne), AI in Healthcare (Harvard Medical School) Senior Surgeon | Governance Leader | HealthTech Co-founder |Harvard Medical School — AI in Healthcare | Australian Institute of Company Directors — GAICD candidate University of Oxford — Sustainable Enterprise Institute for Systems Integrity (ISI) Abstract Healthcare AI governance has entered
This paper examines why systems resist change under pressure and how structurally protected variables shape governance behaviour and outcomes.
The Systems Integrity Toolkit – Phase I introduces a governance architecture for understanding how integrity fails under system stress and how organisations can intervene before harm occurs.
A systems analysis of how urgency compresses judgment, suppresses signals, and accelerates governance failure — and why pause must be designed as a control condition.
This paper introduces the Failure Taxonomy — a structural model showing how harm accumulates in complex systems through drift, signal loss, and accountability inversion, without anyone intending it.
Governance failure rarely begins with misconduct. It begins when constitutions, delegations, decisions, and oversight drift out of alignment under pressure. This paper explains how compliance can persist even as integrity quietly erodes.
Integrity is often treated as a personal trait. This paper shows why it is better understood as a system property — shaped by how authority, accountability, and information are aligned under stress, and why outcomes reflect design rather than intent.
When frontline teams keep systems functioning through heroics and sacrifice, governance has already failed. This ISI paper explains how resilience hides systemic risk.
Governance systems are designed for stability. Under sustained stress, visibility distorts, oversight becomes selectively blind, drift normalises, and outcomes become predictable.
Dr Alwin Tan, MBBS, FRACS, EMBA (University of Melbourne), AI in Healthcare (Harvard Medical School) Institute for Systems Integrity Introduction: the limits of individual explanations When adverse outcomes occur, organisations often focus first on individual actions. Questions such as who made the wrong decision or who failed to follow the
Welcome — Why This Institute Exists This platform exists to explore system integrity — in health, governance, technology, and public decision-making — at a time when trust is under strain and complexity is accelerating. Across institutions worldwide, we are seeing the same patterns repeat: * Decisions made far from those who carry their consequences
Welcome — Why This Institute Exists This platform exists to explore system integrity — in health, governance, technology, and public decision-making — at a time when trust is under strain and complexity is accelerating. Across institutions worldwide, we are seeing the same patterns repeat: * Decisions made far from those who carry their consequences