Authentic Intelligence in Digital Strategy Systems: A Socio-Technical Analysis of Human-Accountable Decision Governance

Background: Digital strategy increasingly relies on algorithmic decision systems, yet the mechanisms by which human judgement remains embedded within these systems are poorly theorised. Existing frameworks treat digital tools as either neutral instruments or autonomous agents, overlooking the systems-level conditions under which human accountability is maintained. Methods: This study employs a novel three-stage system-oriented analytical protocol: (1) mechanism-revealing thematic analysis of 50 semi-structured interviews with senior managers across multinational organisations; (2) configurational cross-case mapping against 685 cases from the European Commission’s JRC AI implementation catalogue; and (3) failure mode triangulation comparing interview-reported barriers with 37 documented implementation discontinuations. Results: We introduce Authentic Intelligence as a systems-level construct and develop a socio-technical architecture specifying six primary system functions, three decision loci, four governance mechanisms, and twelve empirically derived failure modes. Triangulation reveals high correspondence (≥20% JRC citation rate) for six failure modes and moderate correspondence for six additional modes. Conclusions: The contribution is a reusable systems architecture and diagnostic framework for maintaining human-accountable decision governance in digital strategy implementation, with direct application to EU AI Act Article 14 compliance requirements.