Boundary Conditions, Tradeoffs, and Failure Modes

The theory applies most strongly when organizations face meaningful AI-agent participation, distributed authority, nontrivial decision impact, dynamic topology, and pressure for adaptation. It is less central when AI systems are used only as low-risk individual productivity tools or when organizational action remains simple, reversible, and human-contained.

AI and technical conditions include AI autonomy, agent density, technical reliability, explainability, and integration depth. Higher autonomy and greater agent density increase the importance of authority provenance, observability, and human governance. Low autonomy, low-risk, and fully reversible uses weaken the need for full constitutional architecture.

Task and decision conditions include decision impact, ambiguity, reversibility, interdependence, and time sensitivity. High-impact, ambiguous, irreversible, or cross-boundary decisions require stronger delegation, evidence, and escalation. Routine, reversible, and low-impact actions may be governed through lighter constitutional provisions.

Organizational conditions include complexity, maturity, governance competence, representation quality, human expertise, and existing trust. Constitutional architecture depends on the organization's ability to maintain representations, interpret evidence, and exercise judgment. Low-maturity organizations may experience constitutional architecture as overhead unless the architecture is staged and proportionate.

Institutional and environmental conditions include regulatory intensity, professional norms, environmental volatility, cultural expectations, and cross-jurisdictional constraints. Regulation may increase the need for formalization and evidence. Volatility may increase the need for adaptation, but also the risk of rushed constitutional change.

Power and contestation are also boundary conditions. Constitutional architecture does not remove conflict over goals, roles, authority, or resource allocation. It makes conflict more governable when actors can contest decisions, appeal delegations, challenge evidence, and identify which authority source is being invoked. It may fail when powerful actors use constitutional provisions to protect local interests, when affected actors lack standing to contest action, or when formal compliance substitutes for substantive legitimacy.

The theory also contains tradeoffs. Autonomy can increase speed but weaken accountability if delegation is opaque. Formalization can increase governability but reduce flexibility if provisions become too rigid. Machine-readable rules can support consistency but may lose contextual meaning. Local autonomy can improve adaptation but fragment enterprise coherence. Transparency can support oversight but create interpretive overload. Resilience can improve repair but reduce efficiency if every action is burdened with excessive controls.

Failure modes follow from these tradeoffs. Constitutional rigidity occurs when provisions cannot adapt to changed conditions. Governance overhead emerges when review costs overwhelm organizational learning. False assurance occurs when visible controls create confidence without meaningful observability. Accountability diffusion occurs when many actors and agents participate but no authority chain remains traceable. Constitutional drift occurs when practice gradually departs from constitutional commitments. Constitutional capture occurs when powerful actors use the constitution to protect local interests. Constitutional paralysis occurs when amendment or exception pathways are too costly. Uncontrolled agent cascades occur when AI agents trigger other agents without bounded delegation. Permanent emergency authority occurs when temporary exceptions become normalized. Compliance theater occurs when organizations maintain formal records without substantive legitimacy. Unobservable delegation occurs when authority is technically exercised but institutionally invisible.

Repair requires detection, authority suspension, containment, rollback, remedy, appeal, constitutional review, amendment, historical preservation, and learning. Constitutional resilience is therefore not the absence of failure, but the capacity to govern failure.