Decision Trust Zones is a framework developed by Morris Misel to map the boundaries of AI decision-making authority in organisations. It emerged from the Who Decides 2025 research — a global study of how leaders across finance, healthcare, education, government, and professional services are drawing lines on what AI can and cannot decide, and why.

The framework identifies four zones of decision-making, defined by two axes: the reversibility of the decision and the consequence of getting it wrong. Each zone has a different appropriate relationship between human judgment and AI assistance. Understanding which decisions belong in which zone — and designing governance accordingly — is what separates organisations that implement AI well from those that implement it in ways that erode trust and accountability.

The Four Zones

The first zone covers routine, reversible, low-consequence decisions. These are the decisions where AI authority is both appropriate and efficient: scheduling, data categorisation, standard query resolution, pattern recognition in large datasets. Human oversight here is light-touch, and AI autonomy is well-suited.

The second zone covers complex, reversible, moderate-consequence decisions. These require AI augmentation rather than AI authority. AI prepares, analyses, and recommends. Humans review, contextualise, and decide. The decision is recoverable if wrong, but the stakes warrant human judgment in the loop.

The third zone covers high-consequence, harder-to-reverse decisions. Credit decisions above certain thresholds, clinical treatment plans, employment actions, significant resource allocations. Here, AI informs but does not decide. Human accountability is explicit, and the decision record must show who made the call and on what basis.

The fourth zone is what Morris Misel calls Humanity Rules: decisions that carry such weight — in consequence, in accountability, in ethical significance — that AI authority is categorically inappropriate regardless of capability. End-of-life care decisions. Criminal sentencing. Strategic military judgments. Constitutional or policy decisions that determine the structure of public life. These decisions are not merely outside AI’s competence. They are outside AI’s legitimate authority, and maintaining that distinction matters for how societies function.

How Decision Trust Zones Is Applied

In practice, Decision Trust Zones gives boards and executive teams a structured way to conduct AI governance reviews. Rather than asking “which decisions can AI make?” — a capability question — it asks “which decisions should AI make?” — a values and accountability question. The distinction matters enormously when building governance frameworks that will hold under public scrutiny and regulatory pressure.

For leaders navigating AI strategy in finance, healthcare, government, or professional services, Decision Trust Zones provides the vocabulary and the structure to have the governance conversations that technical AI frameworks alone cannot support.

Morris Misel delivers keynotes and workshops on Decision Trust Zones for boards, associations, and leadership teams. More at morrismisel.com/framework.

Decision Trust Zones and the Who Decides 2025 Research

The Who Decides 2025 research found that leaders in every sector had formed intuitions about these zones — they knew, instinctively, which decisions felt appropriate for AI and which did not. What they lacked was a framework for articulating and governing those intuitions at the organisational level. Decision Trust Zones provides that structure.

The research also found that organisations with explicit, documented decision governance boundaries experienced significantly higher stakeholder trust in their AI implementations than those that allowed AI authority to expand organically without deliberate design.