Since 2015, China has witnessed a rapid increase in new energy vehicle (NEV) market penetration, achieving global leadership in this sector. This study employs social network analysis (SNA) and Granger causality tests to examine how policy coordination has influenced China’s NEV market development from 2015 to 2023. We evaluated policy coordination using six network metrics: network density, average path length, transitivity, average clustering coefficient, number of components, and size of largest component. Our findings reveal both correlative and causal relationships between policy coordination and market performance. The analysis demonstrated strong positive correlations between network metrics and market performance indicators (ρ = 0.800–0.850, p < 0.01), while Granger causality tests identified significant temporal effects, particularly in the long term (F = 284.051–281,486.748, p < 0.001). Notably, the largest component size shows immediate causal effects on market performance (F = 4.152, p < 0.05). Based on these results, we recommend establishing a multi-level policy coordination system, optimizing the policy network structure with emphasis on core components, implementing dynamic policy adjustment mechanisms considering time-lagged effects, and strengthening collaborative supervision of policy implementation to further advance China’s NEV market development.

With the maturity and proliferation of large-scale artificial intelligence language models, it has become challenging to prevent university students from using chatbots like ChatGPT as auxiliary tools for course learning. An increasing number of higher education institutions are adopting a positive attitude towards AI language models and encouraging students to use various chatbots. Based on the Technology Acceptance Model (TAM), this study constructs a new theoretical model to explore the influence of perceived organizational support (POS) on students' intention to use AI language models in course learning. Using questionnaire data from 146 university students, this study validates the hypothesized model using PLS-SEM. We found: firstly, TAM has good explanatory power for the intention to use AI language models among university students, with Perceived Usefulness (PU) and Perceived Ease of Use (PEU) both having a positive and significant impact on Behavioral Intention (BI); secondly, although POS cannot directly affect BI, it can positively and effectively influence BI through the mediating roles of PU and PEU. The model has a high explanatory power (R2=0.697), providing valuable theoretical guidance for universities to support and regulate the use of AI language models in course learning by students.

Whether China’s carbon trading pilots stimulate firm-level innovation remains contested. This study investigates the conditional effect of carbon pricing on green invention patents, moderated by firms’ digital transformation capability. Using an unbalanced panel of 31,048 firm-year observations from A-share listed companies (2011–2022) and a two-way fixed-effects difference-in-differences estimator, we find no significant unconditional average treatment effect, consistent with attenuated price signals in China’s pilots. However, digital capability substantially amplifies the innovation response: a one-standard-deviation increase in within-year digital capability is associated with a 4.6% rise in expected green invention patents among treated firms. The moderating effect is exclusive to examination-intensive invention patents rather than utility model registrations, consistent with a strong-form Porter mechanism. Sub-dimension analysis identifies strategic leadership and digital achievement as the operative channels. The results are robust to alternative estimators, cohort decomposition, and sample restrictions. Together, the findings reframe digital transformation as a policy-contingent complementary asset and extend the Porter hypothesis to an organizational-capabilities dimension.

Connected and Automated Electric Vehicles (CAEVs) are rapidly evolving into complex Cyber-Physical-Social Systems (CPSS), generating structural tensions between technological innovation and public safety. Current research in public governance exhibits significant fragmentation. Scholars frequently isolate data privacy compliance from algorithmic safety auditing, treating them as distinct silos. To bridge this gap, this study applies the PRISMA framework to systematically synthesize 135 core peer-reviewed articles, exposing the endogenous limitations of unidimensional regulatory paradigms. Our analysis yields three central insights. First, traditional “notice-and-consent” models fail under the ubiquitous data collection demands of modern V2X environments. Macro-level policies must translate into foundational Privacy-Enhancing Technologies (PETs) through “Law-as-Code” mechanisms. Second, the opacity of end-to-end algorithmic decision-making deconstructs traditional tort liability systems. This necessitates ex-ante quantitative auditing mechanisms—such as Explainable Artificial Intelligence (XAI) and enhanced Threat Analysis and Risk Assessment (TARA 2.0)—to mitigate adversarial attacks and physical-level safety hazards. Third, overcoming cross-national regulatory fragmentation requires constructing a “dual-track synergistic” governance architecture. This framework institutionalizes the coupling of data lifecycle quality workflows with the algorithmic Safety of the Intended Functionality (SOTIF). Ultimately, this review advocates for adaptive regulatory sandboxes and advances the harmonization and mutual recognition of global standards (e.g., ISO/SAE 21434, UN R155/156). Addressing current methodological and empirical data constraints, future academic inquiry must pivot. Researchers should target the value alignment challenges of Large Language Models (LLMs) in autonomous driving and implement multi-stakeholder participatory policy pilots designed to reconcile diverse social values.

The energy implications of transport electrification are presumed to fall on the power grid, yet the fuel and electricity margins are rarely examined in parallel, and never in a tourism-driven economy heavily dependent on imported electricity. This paper takes Macau—where approximately ninety percent of electricity is imported, supply reliability is saturated, and tourism dominates demand—as an extreme case. Using monthly data from January 2010 to July 2026, we estimate within a nonlinear autoregressive distributed lag (NARDL) framework the long-run equilibrium associations between electric vehicle (EV) penetration and two distinct dimensions, namely gasoline consumption (fuel substitution margin) and electricity import dependence (electricity margin). Bootstrap bounds testing accommodates the near-I(2) character of the penetration series. The fuel margin shows a negative long-run association, with log-form point estimates spanning −0.27 to −0.62. The association is statistically firm in the full and handover windows, whereas the EV-era estimate halves to −0.27 and loses conventional significance once residual seasonality is corrected. Identification is contingent on the diffusion trend, so the range is interpreted as an upper bound on substitution rather than a magnitude attributable to electrification alone. The electricity margin shows a bounded null: eight of nine specifications yield no detectable association; the sole significant estimate is form-dependent and inconsistent with observed trends. Minimum detectable effect calculations and the observation that charging load constitutes less than 0.4% of total electricity consumption jointly bound any plausible effect. Tourism exhibits asymmetry consistent with a ratchet in the full and handover windows, where the elasticity with respect to visitor inflows (0.08–0.09) exceeds that of outflows (0.05–0.07), with the pattern attenuating under alternative specifications. In this constraints-driven tourism economy, electrification is reshaping the fuel ledger while the electricity ledger remains undisturbed.

Chinese provinces combat forest fires and pest outbreaks largely by issuing policy documents, yet whether document strength corresponds to lower disaster losses—and which instruments work—remains untested at the province level. We coded the full text of 4503 fire prevention and pest control policies (estimation window 1998–2024), scoring issuing authority, genre force, and instrument strength, and aggregated scores into province-year policy stocks by dominant type: coercive mandates, incentives, and capacity-building. Linking these stocks to forest-masked MODIS active fires and yearbook pest records in two-way fixed-effects models, we found no average association between overall strength and losses. The null masked sharp type differences: a one-standard-deviation increase in incentive strength was associated with about 7% fewer fire points, robust under wild cluster bootstrap inference and Monte Carlo simulation—although not after multiple-comparison correction across the pre-specified family; capacity-building showed weaker, inference-sensitive associations; and coercive mandates—equaling three quarters of the documents—showed none. Associations emerged with a one-year lag and tracked sustained accumulation rather than issuance. For pest losses, results were directionally consistent but weaker. After the forest chief system rollout in 2021, the overall association turned significantly negative. These results help explain why aggregate policy measures often miss policy effects.