Artificial intelligence as a collective transformative experience: a conceptual framework for rational encounter

Purpose Artificial intelligence (AI) is transforming social structures, value systems and collective knowledge practices in ways comparable to transformative experiences, which fundamentally alter preferences and perspectives. This paper aims to examine how the conceptual framework of transformative experiences can inform societal responses to AI-driven change. It asks: how can the theory of transformative experiences guide collective decision-making and governance under conditions of epistemic uncertainty and value transformation associated with AI? The aim is to develop a conceptual framework for understanding AI as a collective transformative experience and for supporting more rational, foresight-oriented societal engagement with its impacts. Design/methodology/approach This study adopts a multidisciplinary conceptual approach, integrating insights from transformative experience theory, social epistemology, futures studies and AI governance literature. Firstly, key governance challenges (uncertainty, opacity and value instability) are identified through a focused analytical review. Secondly, structural analogies between individual transformative experiences and societal encounters with AI are developed. Thirdly, strategies for rational engagement with transformative experiences are translated into a phased governance framework comprising preparation, evaluation, implementation and reflection, with explicit alignment to established foresight methods and tools. Findings AI-induced societal change shares core features with transformative experiences, including deep epistemic uncertainty, limited ex ante evaluation and the potential for preference and value transformation. The analysis argues that, under these conditions, governance is likely to depend on collective epistemic processes rather than purely technical optimisation. On this basis, the paper proposes – as a conceptual contribution to be tested empirically – a four-phase governance model (preparation, evaluation, implementation and reflection) organising a representative and extensible taxonomy of 11 strategies, rather than a closed or exhaustive set, to support adaptive, participatory and learning-oriented governance. Integrating social epistemology foregrounds the potential importance of epistemic diversity, institutional learning and structured collective deliberation for managing AI-related transformation. The framework is stated in falsifiable terms, and its own operationalisation is offered for empirical scrutiny. Research limitations/implications The proposed framework is conceptual and requires empirical validation. This paper specifies explicit conditions under which the framework would be disconfirmed and identifies concrete empirical tests. Future research should test and refine the model through case-based studies of AI deployment in domains such as health care, public administration or autonomous systems. Empirical work is also needed to examine how epistemic institutions and participatory mechanisms influence collective learning under transformative technological conditions. Practical implications The framework offers guidance for policymakers and organisations seeking to structure decision processes under deep uncertainty. Recommended practices include anticipatory governance, experimental regulatory environments, multi-stakeholder deliberation, iterative evaluation and foresight-based learning mechanisms to support adaptive institutional responses. Social implications Viewing AI as a collective transformative experience emphasises the need for inclusive and reflexive governance that addresses value change, distributional effects and power asymmetries. Such an approach may help align AI development with evolving societal priorities related to fairness, accountability and human well-being. Originality/value This paper reconceptualises AI governance through the lens of collective transformative experience and integrates this perspective with social epistemology. By developing a phased model organising a representative taxonomy of strategies grounded in epistemic theory and aligned with foresight practices, it offers a novel socio-epistemic framework that extends beyond risk management and technical ethics towards adaptive, learning-oriented and futures-aware governance. The framework is presented reflexively and in testable form, inviting empirical evaluation of both its prescriptions and its own operationalisation.