With the rapid progression of urbanization, Chinese cities are encountering an increasingly severe shortage of land resources. To examine the role of digital finance (DF) in promoting economic agglomeration (EA) and subsequently improving urban land use efficiency (ULUE), this paper focuses on 274 prefecture-level cities from 2013 to 2020. By employing an interaction effect model, the study analyzes how DF influences the relationship between EA and the enhancement of ULUE. The findings demonstrate that DF positively moderates the effect of EA on enhancing ULUE. The analysis reveals that DF can accelerate resource flows and more effectively promote the enhancement of ULUE by amplifying the positive influence of EA on regional innovation ability (IA) and the rationalization of industrial structures (RIS). Furthermore, the heterogeneity analysis indicates that this moderating effect is particularly pronounced in China’s central and eastern regions, areas with higher rates of internet penetration, and regions with stronger innovation capacities. Based on these findings, this paper proposes policy recommendations such as promoting regional integration and supporting innovation in DF. These insights contribute to the theoretical discourse on EA in the digital era and provide critical and practical guidance for optimizing the development of agglomeration economies and constructing a high-quality development framework.

Against the background of global climate change and increasing ecological vulnerability, enhancing ecosystem resilience has become a core task for coping with environmental shocks and achieving sustainable development. The urban energy structure plays a critical role in influencing the green development of the economy and the enhancement of environmental resilience. Existing studies have revealed the role of energy structure transformation in the identification of macroeconomic performance and environmental outcomes, but have neglected its impact on ecosystem resilience. This paper exploits the implementation of the New Energy Demonstration City pilot policy as a quasi-natural experiment. Using panel data of Chinese prefecture-level cities from 2010 to 2022, it constructs a multidimensional evaluation system of urban ecosystem resilience and employs a difference-in-differences (DID) model to empirically examine the impact of energy structure transformation on urban ecosystem resilience. It is found that energy structure transition significantly enhances urban ecosystem resilience, and this conclusion is verified through a series of robustness tests. Mechanism analysis shows that energy structure transformation comprehensively enhances urban ecosystem resilience through strengthening institutional regulation, optimizing resource allocation, promoting energy substitution, and enhancing public awareness. Heterogeneity analysis indicates that the strengthening effect of energy structure transition on urban ecosystem resilience is inclusive, and that this positive effect is greater in cities characterized by lower resource endowment and weaker governance capacity. This paper reveals the intrinsic mechanism of urban energy transition for ecological resilience enhancement, and provides an energy transition path for building more resilient urban ecosystems.

This study examines whether and how generative artificial intelligence (AI) stimulates technological innovation in biopharmaceutical enterprises. We define the generative AI shock using the public release of ChatGPT in November 2022 and an unbalanced panel of 2002 firm-year observations for Chinese A-share listed biopharmaceutical firms from 2017 to 2024. We construct a firm-level pre-shock AI exposure index from 2021 annual reports and combine high versus low exposure with a post indicator for 2022–2024 in a difference-in-differences framework. Results show that firms with greater pre-shock AI exposure experience significantly stronger increases in invention patent output, with the findings remaining robust across alternative specifications, sample restrictions, treatment-timing definitions, and firm-level clustered inference. According to the mechanism analyses, generative AI promotes innovation through stronger resource acquisition, greater specialisation and focus, and increased external collaboration. Heterogeneity tests show that effects vary across resource endowments, organisational capabilities, and innovation conditions. Further analyses reveal improvements in innovation quality and technological search breadth, with stronger marginal gains among firms with weaker accumulated knowledge. These findings show how the release of ChatGPT, as a generative AI shock, can reshape innovation in a knowledge-intensive and high-uncertainty industry.