Artificial intelligence has moved from the margins of creative practice into the very heart of cultural production, and nowhere is that shift more contested than in the realm of intangible cultural heritage—the crafts, performances, rituals, and design traditions that communities have transmitted across generations. A new study published in Scientific Reports examines how audiences actually respond when machine-generated designs draw on intangible cultural heritage, and its findings suggest that the path from visual appeal to genuine cultural acceptance runs through something far harder to engineer: authenticity. Using a stimulus–organism–response framework adapted from environmental psychology, the research maps the psychological journey that transforms an aesthetically pleasing AI-generated heritage design into a design that audiences are willing to embrace, share, and support.
The study, conducted by Feng and Hu, addresses a puzzle that has grown more urgent as generative tools proliferate. Museums, design studios, and cultural entrepreneurs increasingly use AI systems to produce textiles, patterns, packaging, and digital artifacts inspired by traditional motifs and techniques. These outputs can be strikingly beautiful, and beauty alone might be expected to guarantee commercial and cultural success. Yet heritage is not merely decoration. It carries embedded meaning—stories, identities, communal memory, and the trace of human hands. The researchers hypothesized that aesthetic quality, while necessary, is insufficient, and that perceived authenticity operates as the decisive psychological gateway between seeing a design and emotionally investing in it.
To test this idea, the team built their investigation around the stimulus–organism–response model, a well-established framework in consumer and environmental psychology that treats external stimuli as inputs that shape an organism’s internal states, which in turn drive behavioral responses. In this adaptation, the AI-generated intangible cultural heritage design functions as the stimulus. The organism comprises the audience’s internal psychological states—specifically aesthetic appreciation and perceived authenticity. The response encompasses behavioral intentions such as willingness to purchase, share, recommend, or otherwise engage with the design. This structure allowed the researchers to specify not just that audiences respond positively or negatively to AI heritage designs, but through which psychological mechanisms those responses arise.
The methodological design reflects a careful effort to capture audience psychology under realistic conditions. Participants were exposed to AI-generated designs grounded in intangible cultural heritage content and then asked to evaluate them across the model’s dimensions: their aesthetic response to the visual form, their judgment of whether the design felt authentically connected to the heritage tradition it referenced, and their downstream behavioral intentions. By measuring these variables systematically and modeling the relationships among them, the researchers could distinguish the direct effect of aesthetics on engagement from the indirect effect that operates through authenticity. This mediational architecture is the study’s central analytical contribution, and it is what elevates the work beyond simple survey findings about whether people “like” AI art.
The results support a two-stage model of audience acceptance. Aesthetics, the researchers find, does matter—but its influence is largely funneled through authenticity rather than acting independently. When audiences perceive an AI-generated heritage design as beautiful, that perception boosts their sense that the design is a legitimate expression of the tradition. It is this sense of legitimacy, in turn, that drives approach behaviors: the desire to own, share, recommend, or celebrate the work. Perceived authenticity thus emerges as the pivotal organism-level variable, the psychological hinge on which the entire response turns. A design that dazzles the eye but rings hollow culturally fails to cross this hinge, and its behavioral impact collapses accordingly.
This finding carries significant theoretical weight. In the psychology of heritage consumption, authenticity has long been treated as a multidimensional and sometimes elusive construct—distinguished, for example, between the authenticity of an object’s origins and the authenticity of an experience it affords. What the stimulus–organism–response framing adds is a clear causal placement: authenticity is not merely another feature audiences evaluate in parallel with beauty, but a mediating state that converts sensory appreciation into meaningful engagement. The study thereby offers a more precise account of why “pretty but shallow” AI heritage content so often fails, and why some AI-assisted designs succeed in winning over skeptical audiences.
The practical implications extend across several sectors. For designers working with generative tools, the message is that prompt engineering alone cannot deliver cultural resonance. AI systems trained on visual data can reproduce surface stylistic features of a tradition—the color palettes, the symmetry, the characteristic motifs—but they do not inherently encode the narratives, values, and communal practices that make heritage meaningful. The study suggests that successful projects will be those in which AI outputs are embedded within human cultural mediation: consultation with tradition bearers, documentation of provenance, storytelling that connects the design to living practice, and transparent communication about the role of AI in the creative process. Each of these interventions plausibly feeds the audience’s perception of authenticity, and thereby strengthens behavioral engagement.
For custodians of intangible cultural heritage—communities, practitioners, and the institutions that support them—the research offers both reassurance and a warning. The reassurance is that audiences do not appear to reject AI involvement categorically; where designs feel authentic, acceptance follows, suggesting that generative tools can serve as instruments of revitalization rather than solely as threats. The warning is that authenticity is fragile. If AI-generated content floods the market with designs that borrow the look of a tradition without its substance, audiences may learn to distrust the category as a whole, devaluing even thoughtful collaborations. The mediating role of authenticity means that reputational damage travels through the same psychological channel as positive engagement—an asymmetry that heritage communities can ill afford to ignore.
The study also speaks to a broader debate in the science of human–AI creativity. Recent years have produced a rich literature on how labeling affects aesthetic judgment: works labeled as AI-made are often rated lower than identical works labeled as human-made, a bias that appears rooted in assumptions about effort, intentionality, and emotional depth. The present research complicates that picture by suggesting that in the heritage domain, the decisive question is not simply “who made this?” but “does this genuinely belong to the tradition it invokes?” Authenticity, on this account, is partly detachable from authorship. An AI-generated design presented within a credible cultural frame can achieve the psychological standing that audiences require, even as an uncontextualized output from the same model fails. This reframing shifts attention from the artifact in isolation to the ecosystem of meaning that surrounds it.
There are, inevitably, boundaries to what a single study can establish. The stimulus–organism–response model, whatever its strengths, captures a snapshot of audience psychology rather than the full developmental arc of how perceptions of AI heritage content evolve over time. Cultural background, prior familiarity with a tradition, and personal values almost certainly moderate the relationships the study identifies, and future work will need to test the model across diverse heritage domains—from weaving and ceramics to music, dance, and festival arts—and across audiences with varying degrees of connection to the source communities. Longitudinal designs could reveal whether perceived authenticity, once established, is durable or erodes as audiences become more sophisticated about AI capabilities. Experimental manipulations of contextual cues, such as provenance statements or practitioner endorsements, would further clarify which authenticity-building interventions are most effective.
Even so, the study arrives at a formative moment. Generative AI systems are becoming capable of producing heritage-inspired designs at a scale and speed that no human craft tradition can match, and the institutions charged with safeguarding intangible heritage are drafting policies about how, whether, and by whom such tools may be used. Decisions made in the next few years will shape whether AI becomes a partner in cultural transmission or a force that flattens living traditions into extractable visual styles. Research of this kind, by identifying authenticity as the mechanism through which audiences extend or withhold their engagement, gives those policymakers an empirical target: the goal is not to ban the technology or to celebrate it, but to govern the conditions under which its outputs can credibly participate in cultural life.
The deeper contribution of the work may be conceptual. By formalizing the journey “from aesthetics to authenticity,” Feng and Hu have given researchers a testable template for studying human responses to AI-generated cultural content across many domains—the same framework could plausibly illuminate reactions to AI-composed folk music, AI-authored heritage narratives, or AI-reconstructed historical performances. In each case, the model predicts that surface appeal will matter only insofar as it nourishes a felt sense of genuine cultural belonging, and that engagement behaviors follow from that sense rather than from beauty alone. As generative systems grow more powerful, the scarcity that determines the value of AI heritage design may not be computational capacity or aesthetic polish, but the far rarer resource of credible connection to the communities whose heritage is being reimagined. This study offers one of the clearest quantitative portraits yet of how audiences make that judgment—and a reminder that in the cultural sphere, perception of the real may matter as much as the real itself.
Subject of Research: Audience psychological responses to AI-generated intangible cultural heritage designs, examined through a stimulus–organism–response model linking aesthetic appreciation and perceived authenticity to engagement behaviors
Subject of Research: Technology and Engineering
Article Title: From aesthetics to authenticity: a stimulus–organism–response model of audience responses to AI-generated intangible cultural heritage design
Article References: Feng, L., & Hu, W. (2026). From aesthetics to authenticity: a stimulus–organism–response model of audience responses to AI-generated intangible cultural heritage design. Scientific Reports. https://doi.org/10.1038/s41598-026-69396-4
Image Credits: AI Generated
DOI: 10.1038/s41598-026-69396-4
Keywords: AI-generated design, intangible cultural heritage, authenticity, aesthetics, stimulus–organism–response model, audience engagement, generative AI, cultural heritage design, behavioral intention, mediation analysis
Cite Scienmag News
APA MLA Chicago
Denise Maddox. (September 4, 2026). AI-designed heritage: how audiences respond to authentic cultural designs. Scienmag. https://scienmag.com/ai-designed-heritage-how-audiences-respond-to-authentic-cultural-designs/
Denise Maddox. “AI-designed heritage: how audiences respond to authentic cultural designs.” Scienmag, 4 September 2026, https://scienmag.com/ai-designed-heritage-how-audiences-respond-to-authentic-cultural-designs/. Accessed 4 September 2026.
Denise Maddox. “AI-designed heritage: how audiences respond to authentic cultural designs.” Scienmag. September 4, 2026. https://scienmag.com/ai-designed-heritage-how-audiences-respond-to-authentic-cultural-designs/
Copy citation Download RIS
Tags: AI in museum and cultural institution practicesAI-designed cultural heritageAI-generated cultural heritage designsaudience perception of AI artaudience perception of AI in artsauthenticity in AI cultural creationsauthenticity in AI heritage projectschallenges of authenticity in AI-assisted cultural productioncultural acceptance of artificial heritage representationscultural identity and AI designemotional engagement with AI-created cultural artifactsemotional response to AI-generated cultural artifactsethical considerations of AI in cultural heritageimpact of machine-designed traditional motifsinfluence of visual appeal on heritage appreciationintangible cultural heritage preservationmachine-generated traditional designspsychological impact of AI heritage representationspsychological response to AI-driven heritage artrole of authenticity in AI cultural designrole of authenticity in cultural acceptancetraditional motifs in AI-generated textiles and patternsuse of AI in traditional crafts and rituals
