Decoding urban adoption of AI‑driven cabs: a mixed‑method investigation in China

• Provides actionable insights for understanding intention to use driverless cabs. • Trust in technology, safety, and effort expectancy drive adoption intentions. • Used a mixed-method design with qualitative and SEM-ANN quantitative analysis. • Price value and facilitating conditions show minimal impact on user intention. • Offers policy and design strategies to promote AV use in urban Chinese contexts. AI‑powered autonomous taxis promise to redefine urban mobility, yet consumer acceptance hinges on a nuanced interplay of technological, social, economic, and psychological factors. In this study, we employed a two‑phase, mixed‑method design. Phase 1 comprised semi‑structured interviews with 40 Chinese consumers, generating rich thematic insights, such as the critical roles of perceived efficiency, trust in automation, and safety logic, alongside nuanced concerns about infrastructure, cost fairness, and technology anxiety. Phase 2 applied an extended UTAUT2 framework using a hybrid PLS‑SEM and ANN approach (n = 764), quantitatively confirming that effort expectancy, trust in technology, and perceived safety are the strongest predictors of intention to use driverless cabs, while user experience, social validation, regulatory support, environmental commitment, and hedonic motivation also exert significant influence. Although facilitating conditions, price value, and technology anxiety did not attain statistical significance, qualitative narratives revealed their complementary relevance in shaping initial perceptions. Integrating both strands, we advance UTAUT2 by embedding context‑specific constructs, such as institutional confidence and ethical decision logic, into its theoretical fabric. Practically, our findings recommend targeted efforts to streamline the booking interface, enhance transparency through public performance dashboards, and leverage government pilot‑lane endorsements to bolster consumer trust. This research delivers a robust empirical foundation for stakeholders aiming to accelerate the uptake of driverless taxi services and contributes a versatile mixed‑method template for future studies in autonomous mobility.