Table 3

Evidence-to-construct mapping (Table 1 factors against empirical evidence)

TOE sub-factor (from Table 1)Empirical statusCitizen-facing AIInternal administrative AI
Relative advantageCorroborated; category-specificCapacity reallocation (≈60% automated enquiries)Amplification of expert judgement (€898.7 M detected, 82% hit rate)
CompatibilityCorroborated; tighter for internalModerate (interfaces decoupled)High (on-premise legacy; 150-yr data)
ComplexityPartially observed; resurfaces as explainabilityLow observedExplainability burden (15% residual, human-in-the-loop)
Top-management supportCorroboratedDirective No. 8/2024 assigns deliverablesDirective No. 8/2024 assigns deliverables
Culture/absorptive capacityCorroborated and refinedIT–policy integrationIT–policy integration
ResourcesCorroboratedHuman-capital scarcity → reallocationData resources (150-yr DB)
StructureCorroborated with qualificationCentralised propagation of DirectiveCentralised propagation of Directive
Regulatory frameworkCorroboratedStronger explainability salience (GDPR)Stronger (EU AI Act high-risk for rights)
Stakeholder pressuresCorroborated and differentiatedCitizen trust/legitimacy dominantTrade-union/employee-council dominant
External supportPartially corroboratedSecondary to constitutional perimeterSecondary to constitutional perimeter
Demographic/fiscal pressuresCorroborated and refinedAggregate driverAggregate driver

Note(s): DB: Database

or Create an Account

Close Modal
Close Modal