AI-Generated Situated Storytelling for Sustainable Tourism: Cognitive and Affective Pathways to Local Heritage Exploration Intention

Heritage interpretation shapes which places enter public view and with what meanings, linking heritage protection under Sustainable Development Goal 11.4 (SDG 11.4) with public engagement. Yet visitors in historic urban districts concentrate on a few prominent sites, while historical buildings, remains and places dispersed through everyday streets are seldom recognized as heritage. This study draws on the stimulus–organism–response (SOR) framework and cognitive appraisal theory to examine how six perceived attributes of AI-generated situated storytelling relate to historical place meaning (HPM), place affect (PA) and local heritage exploration intention (LHEI). Data from 548 on-site walking experiences in Nanjing were analyzed using structural equation modeling and artificial neural networks. Place contextualization and perceived personalization were associated with both cognitive and affective pathways; historical contextualization and narrative coherence were associated only with HPM, whereas narrative vividness was associated only with PA. HPM and PA showed comparable direct associations with LHEI, and the model-implied indirect association via PA accounted for 22.1% of the total association between HPM and LHEI. In neural-network models, narrative vividness ranked higher in predicting PA than HPM. These findings support attribute–pathway matching and inform the design of AI interpretation intended to encourage willingness to explore dispersed, low-visibility heritage.