When Ordinary People Stop Imagining AI’s Future: Inside China’s Grassroots Governance Divide

Artificial intelligence is often described as a technology that everyone has an opinion about, yet very few people get to shape. A new study published in Humanities and Social Sciences Communications offers an unusually detailed look at what happens when ordinary citizens are asked to articulate those opinions in depth, and the results reveal a striking tension: people hold vivid, competing visions of what AI should become, but many feel profoundly shut out of the process of turning those visions into reality. The research, conducted by Shuo Wang of Shanghai Jiao Tong University and Jilin University, examines how grassroots publics in China construct and contest what scholars call sociotechnical imaginaries of AI governance, and it documents a phenomenon the author terms imaginative fracture, a quiet withdrawal from public discourse when idealized futures feel permanently out of reach.

The concept of sociotechnical imaginaries sits at the heart of the study. Originating in the field of science and technology studies, sociotechnical imaginaries are collectively held, publicly performed visions of desirable futures that are animated by, and expressed through, science and technology. They are not idle daydreams. Imaginaries do real work in the world: they legitimize certain research programs, justify particular regulatory arrangements, and mobilize publics either in support of a technology or against it. National AI strategies, corporate roadmaps, and government white papers all encode imaginaries of the future, and those encoded visions then circulate through media, education, and everyday conversation, shaping what citizens believe is possible, probable, and permissible. When a society debates facial recognition, algorithmic hiring, or automated content moderation, it is ultimately negotiating between rival imaginaries of the good technological life.

What makes Wang’s study methodologically distinctive is its integration of the sociotechnical imaginaries framework with a second analytical tool known as the Technological Frame. Developed originally to explain how different user groups make sense of emerging technologies, the Technological Frame concept directs attention to the shared interpretive resources a group brings to a new artifact: the problem definitions it applies, the solutions it imagines, and the experiences and metaphors it draws upon. By fusing the two frameworks, the study gains a layered analytical instrument. Sociotechnical imaginaries capture the macro-level, collective visions of technological futures, while technological frames capture the micro-level, group-specific sensemaking through which non-expert participants actually build those visions in conversation. The approach allows the researcher to trace how abstract national narratives about AI are translated, distorted, and sometimes rejected at the level of everyday understanding.

The empirical foundation of the study consists of six online focus groups involving thirty non-expert participants. Focus groups are a qualitative method well suited to this kind of inquiry because they surface the reasoning behind opinions rather than merely tallying them. In a group setting, participants challenge one another, refine their arguments, and reveal the cultural raw material from which their views of AI are assembled. The analysis organized participants’ talk around three dimensions of AI imaginaries: social representations, meaning how people collectively picture what AI is and does; governance mechanisms, meaning who they believe should control the technology and how; and future trajectories, meaning where they think the technology is heading and where they believe it ought to head. This three-part structure gives the study a systematic architecture, ensuring that the findings address not only what people think about AI but how they think it should be steered.

The central empirical finding is a marked divergence between two broad families of imaginaries, which the study labels instrumentalist and humanistic. The instrumentalist imaginary frames AI primarily as a tool, a means of efficiency, productivity, and problem-solving whose value is measured by performance. Within this frame, governance questions tend to be framed as technical optimization problems: how to make systems more accurate, more reliable, more useful. The humanistic imaginary, by contrast, frames AI in terms of its consequences for human dignity, autonomy, employment, and social relationships. Within this frame, governance is not an engineering afterthought but the central question, because the point of regulation is to keep the technology subordinate to human ends. These two imaginaries generate competing narratives at the intersection of technological development and governance, and the study shows that they coexist uneasily within the same population, sometimes within the same individual, producing genuine internal contradiction rather than a simple split between AI enthusiasts and AI skeptics.

This divergence matters because imaginaries are not merely descriptive; they are prescriptive. A person who imagines AI as an instrument will tend to accept governance arrangements that emphasize innovation and deployment speed, while a person who imagines AI as a force requiring humanistic constraint will demand accountability, transparency, and participatory oversight. When a public is divided between such frames, it becomes difficult to form the kind of durable consensus that governance institutions rely upon. The study’s findings suggest that the contest between instrumentalist and humanistic visions is not being resolved in public deliberation but is instead left suspended, with each imaginary pulling governance expectations in a different direction. In the absence of channels through which these competing visions can be articulated and reconciled, the disagreement festers as a source of ambivalence rather than functioning as productive democratic debate.

The most consequential finding, however, concerns what the study calls sociotechnical imaginaries fracture. Wang documents that the participants’ imaginaries, however vivid, are constrained by structural power asymmetries that limit grassroots publics’ capacities to envision future technological directions. In practical terms, the people most affected by AI deployment have the least influence over the scenarios in which that deployment is imagined. The consequential futures of AI are drafted by states and corporations, and ordinary citizens encounter those futures as finished products rather than as open questions. The study finds that this asymmetry produces a loss of imaginative agency: when people cannot see a plausible pathway from their own preferred future to any real decision-making process, the idealized future begins to feel inaccessible, and the imaginary itself fractures.

The behavioral consequence of this fracture is retreat. According to the study, individuals withdraw from governance discourse, stepping back from debates they perceive as having no room for their voice. More troubling still, some participants come to interpret AI development not as a shared project but as a vehicle for disempowering control, a technology whose primary function, as they experience it, is to monitor, manage, and constrain them. This perception represents a significant governance risk. Public legitimacy is a necessary condition for any long-term technological regime, and a public that has stopped imagining itself as a stakeholder will not spontaneously grant that legitimacy. The study thus highlights how structural power asymmetries do not merely exclude people from decisions; they corrode the imaginative foundations on which inclusive technological futures depend.

The implications of the research extend well beyond the Chinese context, even though its setting is specific. Sociotechnical imaginaries fracture describes a dynamic that observers of technology policy in many countries will recognize: publics that are consulted through surveys and comment periods but excluded from agenda-setting, and that respond with disengagement or cynicism rather than opposition. The study’s theoretical contribution, integrating imaginaries with technological frames, provides a transferable template for diagnosing this condition elsewhere, because it links collective visions to the interpretive resources and structural positions of the groups that hold them. For governance designers, the message is that participation cannot be limited to ratifying futures that have already been imagined by others. If citizens are to remain engaged with AI governance, they need genuine imaginative agency, the capacity not only to react to proposed futures but to author them.

The study, which received ethical approval from the Institutional Ethics Committee of the School of Journalism and Communication at Jilin University and was published open access on 14 October 2026, arrives at a moment when governments worldwide are racing to build AI regulatory frameworks while public trust in the technology remains fragile. Its core insight is deceptively simple: the battle over artificial intelligence will not be won solely in laboratories, legislatures, or boardrooms, but in the collective imagination of the people who must live inside the futures being built. Where that imagination is supported, competing visions can be negotiated into something resembling a shared project. Where it fractures, the result is silence, and silence, as this research makes clear, is not consent.

Subject of Research: Grassroots public imaginaries of AI governance in China

Article Title: Support and fracture: understanding sociotechnical imaginaries in Chinese grassroots’ AI governance frames

Article References: Support and fracture: understanding sociotechnical imaginaries in Chinese grassroots’ AI governance frames. (n.d.). https://doi.org/10.1038/s41599-026-09356-5

Image Credits: AI Generated

DOI: 10.1038/s41599-026-09356-5

Keywords: artificial intelligence, AI governance, sociotechnical imaginaries, technological frames, public engagement, science and technology studies, China, focus groups, power asymmetries, technology policy, public participation, qualitative research