Governing Wicked Problems in Healthcare: An Integrity Architecture for AI, Sustainability, and Net Zero

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 — Smith School of Enterprise and the Environment (Sustainable Enterprise)

Institute for Systems Integrity (ISI)

Abstract

Healthcare systems are entering a period of compound transition. Artificial intelligence adoption, net zero commitments, and sustainability obligations are converging within institutions that were not originally designed to absorb this level of complexity, pace, or uncertainty. These challenges are frequently framed as technical, financial, or compliance problems. This paper argues that such framing is insufficient.

Healthcare sustainability and AI governance are better understood as wicked problems: problems with no definitive solution, involving multiple actors with competing objectives, evolving risks, and system-wide consequences. Treating them as tame or technical problems produces predictable integrity failures, including fragmented accountability, informal risk absorption, and moral injury at the frontline.

Drawing on systems leadership and governance theory, this paper proposes an integrity-based governance architecture suited to wicked problems. Rather than pursuing elegant solutions or static controls, healthcare systems must adopt deliberately adaptive and “clumsy” governance designs capable of detecting, surfacing, and correcting harm before it becomes normalised.

1. Wicked problems as a governing reality

Wicked problems cannot be solved once and for all. They resist linear planning, cannot be delegated to a single authority, and evolve in response to attempted interventions. They are defined not by technical difficulty, but by contested values, uncertainty, and interdependence.

Healthcare sustainability and AI governance meet all of these criteria.

Efforts to reduce medical waste intersect with infection control requirements. Decarbonising supply chains intersects with cost, access, and resilience. Deploying AI intersects with clinical judgement, accountability, bias, and trust. Protecting workforce wellbeing intersects with productivity pressures and resource constraints.

Progress in one domain often generates pressure or risk in another. These tensions are not implementation errors. They are structural features of the system. Treating them as temporary or resolvable obscures the real governance challenge.

2. The integrity failure mode

Across healthcare systems, the Institute for Systems Integrity observes a recurring pattern when wicked problems are approached through conventional governance models:

• ambitious sustainability, AI, or net zero commitments are articulated at the institutional level;
• implementation responsibility is decentralised to services, departments, or individuals;
• assurance capability varies widely across organisations and regions;
• learning remains siloed and non-portable;
• unresolved risk is absorbed informally by frontline professionals.

The result is formal compliance with uneven integrity. Policies exist, committees are formed, and reporting requirements are met. Yet accountability diffuses, early warning signals are missed, and harm accumulates quietly upstream of dashboards.

This is not a failure of professionalism or intent. It is a failure of governance architecture.

3. Why elegant solutions fail

Healthcare governance has traditionally favoured elegant solutions: clear lines of accountability, fixed standards, and one-off approvals. These approaches work well for stable, well-bounded problems. They perform poorly under conditions of ambiguity, adaptation, and drift.

Wicked problems punish elegance.

AI systems change as data, workflows, and incentives change. Sustainability interventions reshape behaviour in unexpected ways. Net zero constraints introduce trade-offs that cannot be resolved through optimisation alone.

When governance assumes stability, systems drift out of alignment faster than oversight can respond. Integrity failures emerge not through rule-breaking, but through normalised workarounds.

4. Leadership for wicked problems

Wicked problems cannot be governed through command, control, or compliance alone. They require leadership capacities that are often under-recognised in institutional design.

Leadership in this context is not about having the right answer. It is about the ability to:

• frame the governing question correctly;
• convene diverse expertise and perspectives;
• hold unresolved tensions without premature closure;
• enable feedback, learning, and course correction over time.

This form of leadership must be embedded structurally, not relied upon personally. Expecting individual leaders or clinicians to compensate for structural misalignment is neither safe nor sustainable.

5. Implications for boards and regulators

For boards and regulators, wicked problems demand a reframing of the core governance question.

The question is no longer:

• Is this compliant?
• Has this been approved?
• Did we meet the target?

The governing question becomes:

Can this system detect integrity drift early and correct itself before harm becomes embedded?

This reframing has direct implications for AI oversight, sustainability strategy, and net zero transition. It shifts attention from static assurance to dynamic resilience, from blame allocation to learning capacity, and from performance snapshots to system behaviour over time.

Conclusion

Healthcare’s future will be shaped not by whether institutions adopt AI, commit to net zero, or articulate sustainability goals — but by how well they govern the wicked problems these commitments create.

There are no silver bullets. There are no elegant solutions. Integrity must be designed into systems, not assumed through goodwill, professionalism, or compliance alone.

Governance architectures that can learn faster than they fail are no longer optional. They are essential to the safe, ethical, and sustainable operation of healthcare systems under pressure.

Reference

Rittel, H.W.J. and Webber, M.M. (1973) ‘Dilemmas in a general theory of planning’, Policy Sciences, 4(2), pp. 155–169.
→ Foundational articulation of wicked problems, establishing why certain societal challenges resist definitive solutions and linear planning.

Grint, K. (2005) ‘Problems, problems, problems: The social construction of leadership’, Human Relations, 58(11), pp. 1467–1494.
→ Introduces the distinction between tame, critical, and wicked problems, and argues for leadership approaches suited to complexity rather than command-and-control.

Grint, K. (2010) Leadership: A very short introduction. Oxford: Oxford University Press.
→ Expands wicked problem theory into leadership practice, emphasising question-led, adaptive responses and “clumsy” solutions.

World Health Organization (2023) Global analysis of health care waste in the context of climate change. Geneva: World Health Organization.
→ Documents the scale of healthcare waste, including reliance on single-use medical products and associated environmental trade-offs.

Emanuel, E.J., Wachter, R.M., McClellan, M.B. and Jamieson, K.H. (2020) ‘The ethics of artificial intelligence in health care’, New England Journal of Medicine, 382(23), pp. 2191–2193.
→ Frames AI in healthcare as an ethical and governance challenge rather than a purely technical solution.

Organisation for Economic Co-operation and Development (OECD) (2021) Artificial intelligence, accountability and governance. Paris: OECD Publishing.
→ Examines how AI shifts accountability, accelerates risk, and challenges existing governance structures.

Smith School of Enterprise and the Environment (SSEE) (n.d.) Systems transitions and sustainability. University of Oxford.
→ Provides the systems-based framing used in the article, including the balancing of human, natural, financial, and institutional capital in complex transitions.

#WickedProblems #SystemsIntegrity #HealthcareGovernance #AIGovernance #EthicalAI #AIinHealthcare #NetZeroHealthcare #SustainableHealthcare #ClimateAndHealth #HealthSystems #ClinicalGovernance #HealthPolicy #DigitalHealth #SystemsThinking #PublicValue #SustainabilityLeadership #FutureOfHealthcare #GovernanceUnderPressure #InstitutionalIntegrity #SmithSchoolofEnterpriseandtheEnvironment #UniversityofOxford #InstituteforSystemsIntegrity #MelbourneBusinessSchool #UniversityofMelbourne