Intelligent automation is a core component of Industry 4.0, enabling artificial intelligence (AI) systems to support or execute operational and managerial decisions in real time. In high-risk industrial environments such as mining and metallurgy, real-time decision-making improves efficiency but also raises critical challenges related to trust, explainability, human oversight, and institutional accountability. This study proposes PRIME–INSPECT, a two-layer socio-technical framework designed to support trustworthy AI-driven real-time decision-making. The PRIME (predict, regulate, interpret, mitigate, execute) layer formalizes the operational decision flow, embedding control mechanisms, uncertainty quantification, and explainability into the automation pipeline. The INSPECT (integrity, navigability, supervisory control, policy maturity, ethical compliance, collaboration, trust calibration) layer defines the organizational and governance conditions required for safe deployment. The framework is conceptually developed through a structured literature synthesis and supported by exploratory empirical grounding through stakeholder perceptions from IT and top management participants, alongside an illustrative industrial use case intended to demonstrate conceptual applicability rather than engineering performance validation. The findings highlight the importance of aligning operational AI processes with institutional safeguards to support calibrated trust and responsible automation. The empirical component is intended to provide conceptual and organizational grounding of framework dimensions rather than quantitative validation of predictive performance. PRIME–INSPECT provides a structured architecture for designing and governing AI-enabled real-time decision systems in high-risk industrial contexts.

AI-native decision systems — in which generative artificial intelligence generates, executes, and adapts decisions in real time — are shifting the centre of gravity of analytical effort. Whereas the principal challenge used to be the extraction of insight, it is now becoming the validation of AI outputs while action is still being taken upon them. Existing approaches address this problem only partially: explainable artificial intelligence and governance frameworks treat validation as an ex post layer; reinforcement learning reduces it to a single scalar reward; control theory and the broader cybernetic tradition formalise the closed loop but neglect its sociotechnical constraints; while the PRIME–INSPECT framework establishes a governance foundation, yet leaves real-time validation implicit. This paper proposes a validation-centric architecture that links the GAVA decision loop (Generate–Act–Validate– Adapt), introduced here, with the PRIME–INSPECT framework and elevates validation to a central analytical function comprising four subdimensions: statistical, operational, cognitive, and governance-related. The proposed architecture is empirically examined on a dual sample of IT professionals and top-management representatives, using descriptive statistics, multiple regression, mediation, moderation, and structural equation modelling. The results confirm the central role of trust, the negative effect of perceived risk, and the importance of top management support, while robustness checks separated by the IT and TMT subsamples further strengthen the structural model. Taken as a whole, the findings support the view that validation should be treated as a first-order organisational stage in AI-native decision systems.