Decision Systems Engineering

The Institute's candidate discipline for decision system design

Decision Systems Engineering is the Institute's proposed discipline for specifying, analysing, and governing the conditions under which consequential authority is exercised in organisations. DSE 1.0 Candidate was established in June 2026. It is now in its Application Phase, being tested against real organisational conditions.

01 Object of Study

The decision is the correct unit of analysis. Not the process that produces it. Not the strategy it expresses. Not the capability that enables it. The decision, the moment at which authority is exercised and accountability is incurred.

DSE's object of study is the decision: specifically, the conditions under which consequential authority is exercised within organisations, and under which that exercise can be observed, evaluated, and made to answer.

Every major governance discipline, Enterprise Architecture, Organisation Design, AI Governance, GRC, Process Engineering, assumes a constitutional layer it does not supply. A layer that specifies who has authority over the decisions those frameworks govern, and under what conditions that authority can be exercised, observed, and made to answer. DSE is the Institute's candidate for providing that layer.

The organisations that will consistently produce superior outcomes in the age of AI are not those with the best AI tools. They are those with the best decision systems.

AI participation in organisations primarily operates at the decision level: not after the decision is made, but before it, through the shaping of the possibility space from which the decision maker chooses. Governance that does not reach the decision does not reach where AI's influence operates.

Observable

The decision system can be seen

The authority structure and its exercise can be seen by governance bodies and those affected. A decision system that cannot be seen cannot be governed.

Evaluable

The decision system can be assessed

The exercise of authority can be assessed against applicable standards. The decision system is reason-responsive, capable of being judged against the conditions that should govern it.

Answerable

The decision system is accountable

Those who exercise authority can be held to account. There is an identifiable agent who can be required to give an account of the decision and its consequences.

02 The Decision Boundary

The Decision Boundary is the line in every consequential decision where AI influence ends and human authority must begin. Specifying this boundary, making it visible and holding it in place, is the most consequential act of decision system design.

The Decision Boundary is not a technical concept. It is a constitutional one. It describes the threshold at which responsibility shifts from the systems that shaped the decision to the authority holder who makes it. Where it is unspecified, authority does not disappear, it migrates to whoever most effectively shapes the conditions of the decision. Without specification, that defaults to AI.

Every AI investment that shapes the possibility space of consequential decisions without governance specification is, the Institute proposes, a drift event. Individually, each appears routine. Cumulatively, they represent a constitutional change that no one authorised. The Constitutional Gap, the divergence between who formally holds authority and who effectively exercises it, is DSE's central analytical construct. Authority drift widens it.

A decision system that is not designed will still exist. It will simply be less observable, less evaluable, and less answerable than one that was.

03 Five Constitutional Properties

The Institute proposes five constitutional properties a well governed decision system must satisfy. These are advanced as necessary conditions, not design preferences. A decision system that fails on any one is not fully governed, regardless of how well it satisfies the others.

I, Authority

Every consequential decision requires an authority structure: a defined allocation of who can decide what. Where authority is unspecified, it migrates to whoever most effectively shapes the conditions of the decision. Authority specification is the first act of decision system design.

II, Possibility

The conditions under which authority can be exercised are bounded by technical, organisational, legal, and cognitive possibility spaces. Decision system design must map what is possible before prescribing what should obtain. DSE does not design idealised authority structures that cannot exist.

III, Accountability

Every consequential decision requires a designated accountable authority, an identifiable agent who can be required to give an account of the decision and its consequences. Where full causal traceability is impossible, the requirement for designated accountability remains.

IV, Observability

A decision system that cannot be observed cannot be governed. Observability is the first logical prerequisite of governance, not a design preference. This applies at every layer: the formal authority structure, the operative structure, and the AI participation layer where the possibility space is constituted.

V, Legitimacy

Decisions satisfy the legitimacy condition when made by a recognised authority through a process meeting standards of transparency, deliberativeness, and procedural adequacy. Both conditions are necessary. A decision made by the right authority through an opaque process may be formally authorised but substantively illegitimate.

04 The Framework

DSE 1.0 Candidate is organised around six provisional laws and four governing axioms. The laws are conjectures: claims about how authority behaves in AI augmented organisations. The Application Phase exists to expose those claims to evidence, contradiction and revision.

Six Provisional Laws

Law I Constitution, Every organisation has a constitutional layer specifying formal authority, accountability, and decision rights. Its divergence from the operative authority structure is the Constitutional Gap.
Law II Displacement, Authority displacement occurs when effective decisional control migrates from formal authority holders. AI investment without constitutional specification is proposed as a primary driver.
Law III Compression, Decision systems compress under pressure, time, cognitive load, ambiguity, narrowing conditions for observable, evaluable, accountable authority. AI participation is predicted to intensify compression.
Law IV Drift, Decision systems drift from their constitutional specification unless actively maintained. The Institute proposes that Constitutional Drift Rate is positive for any unmaintained system.
Law V Legitimacy, Decision legitimacy depends on alignment between exercising and recognised authority, and on the quality of the process. Both conditions are necessary. Neither is sufficient alone.
Law VI Observability, A decision system that cannot be observed cannot be governed. Observability is the first logical prerequisite of governance, not a design preference.

Four Governing Axioms

Axiom 1 Authority, Every consequential decision requires a defined authority structure. Authority is structural, not incidental. This axiom is the constitutional starting point for the discipline.
Axiom 2 Possibility, Authority conditions are bounded by what is technically, organisationally, legally, and cognitively possible. DSE maps possibility before prescribing what should obtain.
Axiom 3 Accountability, Every consequential decision requires a designated accountable authority. Where causal traceability is impossible, the accountability requirement remains and must itself be governed.
Axiom 4 Legitimacy, Decisions are legitimate when made by recognised authority through adequate process. Both conditions are necessary. Neither is sufficient alone.

05 Positioning

Every major governance discipline assumes a constitutional layer it does not supply. The Institute proposes DSE as the discipline that provides that layer, not as a competitor to existing frameworks, but as the foundation upon which they depend.

Discipline
Governs
Assumes (but does not specify)
Enterprise Architecture
Technology systems enabling decisions
That authority holders exist with specified decision rights over what the technology enables
Organisation Design
Reporting structures and formal role responsibilities
That roles carry specified decision rights over categories of consequential choice
AI Governance
Risk properties of AI models and compliance of outputs
That human authority holders exist with power and accountability to act on AI outputs
GRC
Compliance with policies, regulations, and procedural requirements
That the policies enforced rest on a constitutional foundation specifying legitimate authority
Process Engineering
Workflows through which decisions are implemented
That the decisions those processes implement were made by legitimate authority holders
Decision Systems Engineering
The conditions under which consequential authority is exercised, observed, evaluated, and made to answer
Nothing. The Institute proposes it provides what the other disciplines assume.
Research Programme Applied DSE The Challenge

06 Lineage

Every era produced frameworks.
None engineered the authority structure.

Pre-1900s 1910s onwards 1950s onwards 1970s onwards 1990s onwards The gap
Command & Hierarchy Behavioural & Organisational Science Structural Frameworks Regulatory & Governance Process, Risk & Technical Decision Systems Engineering
Military hierarchy, imperial administration, guild structures Taylor, Fayol, Weber, Drucker, Mintzberg RACI, RASCI, org design, Enterprise Architecture, Zachman, Holacracy Corporate governance, agency theory, Sarbanes-Oxley, Basel, ISO 38500 COBIT, ITIL, Lean, Agile, enterprise risk, AI ethics, EU AI Act Conclusa Research Institute
Authority personal, positional, traditional. Enforced through rank. Optimised how people decide. Codified roles and hierarchy. Mapped who does what. Described authority flows. Accountability at ownership and compliance boundaries. Optimised outcomes, process and tooling. Authority remained assumed. Treating consequential authority as an engineering object.
Assumed. Not designed. Not the architecture they decide within. Described. Not engineered. Boundaries. Not internal authority flow. Outcomes and tools. Not authority structure. Designed. Observed. Measured.

Decision Systems Engineering treats consequential authority as an engineering object: designed, observed, measured and held.