Socio-Affective Governance in AI: Exploring the Possibility of a Governance Vacuum

This preprint proposes a conceptual and exploratory framework for examining whether generative AI ecosystems may, under certain conditions, exhibit what the paper terms a socio-affective governance vacuum. The concept is introduced as a hypothesis and analytic lens, not as a diagnosis or claim about existing systems, organizations, or communities. Drawing on established research in emotional labor, open digital governance, online moderation, and science and technology studies, the paper synthesizes patterns observed in adjacent sociotechnical domains and asks whether similar dynamics might plausibly emerge in AI contexts where systems intersect with emotionally or culturally meaningful practices. The analysis is deliberately non-prescriptive and non-evaluative, and does not advance policy recommendations, governance models, or institutional critiques. The paper’s contribution is agenda-setting and descriptive. It clarifies conceptual boundaries between socio-affective governance and other AI governance domains (such as safety, bias, and compliance), articulates conditions under which socio-affective considerations could become analytically relevant, and outlines a research agenda for interdisciplinary inquiry. All claims regarding AI are framed as hypotheses grounded in analogy, not as empirically demonstrated findings. Version 1.1:– Minor updates for link maintenance– Added reference to related descriptive framework for interpretive analysis[Suenrith Nguyen, H. (2026). Socio-Affective Interpretation in AI Discourse: A Descriptive Frame (v1.1). Zenodo. https://doi.org/10.5281/zenodo.18237174] The work is intended to support conceptual clarification and future empirical research rather than adjudicate outcomes or assign responsibility.