Powering Energy Security: The Essential Roles of Green Finance and Artificial Intelligence

Internal Auditing & Risk Management Year XX, No 2(72) September 2025 55 POWERING ENERGY SECURITY: THE ESSENTIAL ROLES OF GREEN FINANCE AND ARTIFICIAL INTELLIGENCE Chi Wei SU, PhD Professor School of Economics, Qingdao University, Qingdao, China [email protected] Meng QIN, PhD Associate Professor School of Marxism, Qingdao University, Qingdao, China [email protected] Abstract: A thorough examination of the essential roles of green finance (GF) and artificial intelligence (AI) is crucial for enhancing energy security (ES). By utilising the TVP-VAR-SV framework, the paper reveals the dynamic interactions among the index of GF, AI, and ES, under the influence of the World Uncertainty Index (WUI). Our findings indicate that in the initial phase, GF has a detrimental influence on ES, but this impact reverses to become beneficial in the subsequent second and third stages, highlighting a notable lag in its positive impact. Similarly, AI’s influence on ES starts off negatively in the early and second stages before turning positive in the third stage, with an even more pronounced delay in its positive impact. GF consistently demonstrates a positive and promotional effect on AI. Conversely, AI’s influence on GF is positive in the first phase but becomes more complex, with alternating positive and negative influences in the subsequent stages. Furthermore, WUI exhibits both positive and negative effects on ES, affirming its validity as a control variable to some extent. Against the backdrop of growing uncertainties in the global energy landscape, this study proposes strategic pathways to enhance the role of GF and AI in strengthening ES. Keywords: energy security, green finance, artificial intelligence, time-dependent 1. Introduction This article attempts to study the nexus of green finance, artificial intelligence, and energy security. In this context, ES refers to the assurance of stable, reliable, and unbroken energy supply to meet societal needs while mitigating risks and uncertainties associated with the energy markets (Lee and Fang,2025; Internal Auditing & Risk Management Year XX, No 2(72) September 2025 56 Usman et al., 2025). Energy insecurity arises from a variety of drivers. Political and geopolitical factors include instability in governance, conflicts between energy-exporting nations, and strategic manipulation. Economic influences such as global economic shifts and fluctuations in the business cycle also contribute significantly (Khan, 2025). Moreover, technological advancements and the transition to new energy sources, particularly renewables and energy storage systems, lay a key role (Bashir et al., 2025; Kim et al., 2025). Natural events like disasters, climate change, and pandemics further compound energy insecurity (Shupler et al., 2021). The new and powerful way to move forward with ES is by means of GF, a revolutionary financial concept that includes green development as an integral part of the financial activities. First of all, GF is able to attract investors to finance projects in the energy field by means of issuing relevant bonds and funds, such as the green bonds of the China Development Bank, which had targeted the financing of renewable energy and energy-saving projects, as pointed out by Shi and Zhao (2023). Secondly, a range of financial instruments offered by GF, including green credits and bonds, significantly accelerates progress in energy technology, enabling energy enterprises to secure essential funding for innovation (Iqbal et al., 2025). Thirdly, GF facilitates the green transformation of the energy sector by financing the green development of energy initiatives (Zhang et al., 2025). It encompasses supporting projects related to clean energy, implementing strategies for emission reduction and energy conservation, as well as initiating programs that are aimed at improving energy efficiency (Wang and Xu, 2025; Zhang and Sun, 2025). Accordingly, GF can act as a powerful instrument in advancing ES. Moreover, AI, including machine learning, deep learning, natural language processing, and computer vision, serves as a powerful tool to enhance ES. The International Energy Agency published a report entitled Digitalization and Energy, which notes that as AI technologies are widely embraced, it is likely to decrease the production cost of oil and gas by 10% to 20%, as cited by Mohammadi and Sohn (2023), Abadi et al. (2025), and Baseer et al. (2025). Furthermore, it is expected to diminish the rates of wind and solar power generation curtailment from 7% to 1.6%. The United States Department of Energy’s ARPA-E division has introduced a strategic plan that emphasizes the utilization of smart design principles for enhancing energy efficiency, reducing emissions, and achieving significant technological advancements. In parallel, ExxonMobil, a prominent international oil and gas corporation, leverages AI to enhance the productivity and success rates of its drilling activities. China’s Alibaba has taken a leading role in the growing digital energy field, concentrating efforts on site energy management, designing energy solutions tailored for data centers, advancing the development of intelligent photovoltaic Internal Auditing & Risk Management Year XX, No 2(72) September 2025 57 systems, and providing power solutions for electric vehicles and modules are among the objectives. To date, Alibaba has delivered digital services to more than 200 power utilities, collaborated with 25 oil and gas companies, and served 20 mining enterprises, while also establishing partnerships with 16 key players in the energy sector. Thus, it’s clear that GF and AI are deeply connected to ES, a vital but under-researched subject. Also, the current studies have not sufficiently explored the dynamic interplay between these three elements, which constitutes the main emphasis of our article’s investigation and discourse. This paper shows three key innovations. Firstly, existing research has mainly focused on isolated relations among GF, energy, AI, and combinations of these factors, neglecting a holistic, global perspective that encompasses all three simultaneously. Accordingly, this research offers a pioneering examination of the complex relationships among GF, AI and ES. Secondly, much of the current research on ES primarily relies on theoretical evaluations and focuses on energy costs and consumption, yet lacks comprehensive empirical data. Thereupon, our study introduces a novel methodology by employing Google Trends data as a precise indicator to gauge the prevalent state of ES. Furthermore, we make a significant contribution by introducing the S& P Green Bond U.S. Dollar Select and the S&P Kensho New Economy RAIC indices as measurements to evaluate developments in GF and AI. Thirdly, the intricate interplay between them is evolves, which has been overlooked in previous studies. This study addresses a significant gap in the literature by employing parameter estimation combined with MCMC to assess the viability of constructing a TVP-VARSV model. The application of this methodology enables a more profound understanding of the intricate relationships among GF, AI and ES, thus adding considerable value to the field. 2. Literature Review 2.1. The relation between GF and energy The existing research has delved into the relation between GF and energy. Madaleno et al. (2022) present the three faces of GF: financing environmental conservation, supporting green technologies, and fostering the development of renewable energy in order to pursue ES. On the other hand, Nepal et al. (2024) go on to offer that GF is able to effectively enhance China’s energy resilience, with particular importance for eastern and western regions in helping industrial transformation so as to develop other economic and social dimensions of ES. Liu et al. (2023) give an indication of the vital role that GF plays in furthering sustainable energy growth, particularly in wind and solar energy. Shi and Zhao (2023) even show some striking results of a dynamic relationship between Internal Auditing & Risk Management Year XX, No 2(72) September 2025 58 GF development and energy security; with one unit rise in GF, there is an initial negative response of ES, followed by a turn for the positive. Variance decomposition results indicate that the trend for the initial shock of GF on ES keeps on increasing with time. According to Bhattacharyya (2023), in the transition from a carbon-intensive, fossil fuel-based system to a sustainable, low-carbon system, much financial supply is needed. The classification framework of GF may act as guiding principles for investors. Finally, this work shows that energy insecurity is an issue which underlines its interaction with GF and ES dynamics, manifestation of the associated uncertainties and risks in supply and demand for energy. According to Lee and Fang (2025), GF exerts a significantly positive effect on the ES of the recipient countries. It especially exerts its impact on the reinforcement of ES more in low-income countries, while in high-income countries, it is not significant. Liu et al. (2025) have cited that GF is a game-changing step toward ES, in those countries or regions where energy transition and economic growth are more profound. Meanwhile, Wang and Xu (2025) emphasize that GF offers robust support for enhancing ES in China. 2.2. The relation between AI and energy Scholars have increasingly concentrated their investigations on examining the connections between AI and energy systems. Ghenai et al. (2022) underscore the growing integration of digitalization and interconnectedness in the global energy sector, fueled by advancements in digital technologies like AI. This digitization allows for the implementation of advanced technologies and strategies, significantly enhancing ES. Yu et al. (2022) emphasize the transformative impact of energy digital twin technology on the process and energy industries, leading to improvements in energy management, service and maintenance, energy-efficient designs, and infrastructure upgrades, while also facilitating seamless integration with local and regional renewable energy resources. Arowoiya et al. (2024) find that various algorithms, including AI and ANNs, are utilized to forecast energy consumption in buildings. According to Huang and Lin (2023), digitalization can markedly lower the carbon intensity of power generation, primarily through energy mix optimization, with this effect strengthening as renewable energy capacity grows. Lu and Li (2024) emphasize that the degree of AI innovation is positively associated with the enhancement of overall energy efficiency, where larger-scale innovations yield more substantial benefits. Additionally, AI innovation strengthens ES by reinforcing environmental accountability and upgrading internal governance systems. Qin et al. (2024) reveal that AI exerts dual effects on renewable energy: while its positive impacts promote renewable adoption, they are unsustainable Internal Auditing & Risk Management Year XX, No 2(72) September 2025 59 long-term due to the cost advantage of non-renewables. Meanwhile, Choi et al. (2024) highlight four key factors influencing AI-driven energy use: energy mix, network infrastructure, trading markets, and industrial structure shifts. Zhao et al. (2025) emphasise that enhancements in ES facilitated by AI are mainly due to improvements in the efficiency of coal and electricity usage. Zhai et al. (2025) find that AI can successfully address inefficient investments in energy enterprises, leading to improved ES. 2.3. The relation between GF and AI In investigating the relation between GF and AI, Liu et al. (2023) confirm that GF functions as a promoter for developments in green technology innovations, which in turn indirectly promotes AI. Xiong and Dai (2023) show that green finance (GF) promotes environmental improvement by advancing AI, where non-state-owned enterprises act as key intermediaries. Zhang et al. (2022) note that lagging technological innovation can restrict the financial sector’s ability to offer diverse energy-saving products, thereby hindering GF development. Nonetheless, as AI advancements reach a more advanced stage, both consumers and businesses would have access to diverse energy-consuming products and innovative machine utilization pathways, which can subsequently stimulate the expansion of GF. Zeng et al. (2024) determine that the beneficial effects of GF on AI are particularly pronounced in eastern and central urban areas, in cities that do not depend on natural resources, in cities with populations under 5 million, and in cities that have relatively modest levels of AI development. Research by Zhou et al. (2025) shows that the issuance of green bonds enhances intelligent manufacturing by an average margin of 1.22%, with green innovation serving as the mechanism through which this impact is achieved. Kuang et al. (2024) indicate that industrial AI promotes the growth of GF in particular segments of the data distribution across various economies. A study by Nepal et al. (2025) reveals that innovations in AI technology propel the development of advancements in GF. 3. Methodology The TVP-VAR-SV model, as described by Dong et al. (2024), integrates features of Time-Varying Parameters, Vector Autoregression, and Stochastic Volatility. A key characteristic of this framework is its time-varying parameters, which enhance its flexibility in reflecting dynamic changes within economic systems. Due to the nature of dynamic relations among economic variables including green finance, artificial intelligence, energy security, and world uncertainty index, this model would be able to capture their time variation with more accuracy by virtue of the time-varying attribute; such was also asserted by Internal Auditing & Risk Management Year XX, No 2(72) September 2025 60 Zhong et al. (2023). VAR accounts for interrelations between different economic variables. Variables in an economic system are usually interdependent; changes in one variable may lead to changes in the other. The VAR aspect allows for the consideration of all the interactions within the variables in a wide scope, hence arriving at more realistic predictions and analyses, according to Qin et al. (2023). Thirdly, the SV component incorporates the random fluctuations that variables generally take in economics. The variability in most economic systems has to take stochastic variations in light of stochastic influences of many stochastic factors. In that perspective, Sevillano et al. (2024) argued that the salient features of the TVP-VAR-SV model will be able to represent those random fluctuations. In a word, TVP-VAR-SV considers both dynamic changes in all aspects of the economic system and inner uncertainty simultaneously; it has excellent practical value in economic analysis and forecast. The TVP-VAR-SV model is formulated as follows: (1) In this context, drawing upon Primiceri’s (2005) framework, the TVPVAR-SV is formulated in such a manner that specific components, , and denoted by specific symbols, exhibit time-varying property. The dynamic parameters in the TVP-VAR-SV model could be written as: where , and . in the Equation (2) would be further rewritten as . Moreover, the study implements the Markov Chain Monte Carlo (MCMC) method within a Bayesian statistical paradigm to perform quantitative assessments, which enables the extraction of conclusions that are both more resilient and precise, as detailed by Zhu et al. (2024). This methodological approach is employed to mitigate errors stemming from the instability of initial values, as advocated by Ha et al. (2024). Furthermore, Gibbs sampling from an MCMC framework has been utilized during this study for the purpose of accurate estimation. We calculate the dynamic impulse responses to GF, AI, ES, and WUI. In doing so, it adopts two methods: the local projection approach proposed by Canova and Gambetti Internal Auditing & Risk Management Year XX, No 2(72) September 2025 61 in 2009, and another approach called generalized impulse response function technique developed by Nakajima et al. (2011). Since most literature compares the impact magnitude by applying the generalized impulse response function method, which is supposed to mitigate estimation biases, such as Qiao et al. (2023), Maghyereh et al. (2024), Rodriguez et al. (2024), and Choi and Hadad (2025), this paper follows suit and applies it to derive the time-varying impulse responses of GF, AI, ES, and WUI. 4. Data In this paper, monthly data from January 2015 to December 2024 is employed in considering the changing relationships among GF, AI, and ES. Research of Global Energy Security starts in 2015 because in this year, the Global Energy Sector faced unprecedented and profound changes. On one hand, the rapid evolution of emerging energy technologies, including renewable energy and shale gas, is progressively altering the conventional energy mix and presenting novel avenues for energy supply. On the other hand, shifts in the international political and economic landscape, particularly the escalation of geopolitical tensions, pose challenges to the stability of energy supply. Concurrently, the escalating issue of global climate change, coupled with heightened environmental awareness, has prompted heightened focus on the safe, clean, and sustainable utilization of energy. Consequently, since 2015, the investigation of global energy security has assumed paramount importance, aimed at delving into the challenges and opportunities associated with energy security and providing a scientific rationale for formulating effective energy policies and strategies. In this study, we select Google Trends as the primary data source, with its reference value being manifested in the global search volume for the keyword “energy security”. A higher ES score indicates a greater risk to energy security, while a lower score means better security. Amid the current wave of technological transformation, GF and AI have become central drivers in advancing ES. In this study, the Global Green Bond Index, sourced from S&P Global, serves as a core measure of GF development. Launched on June 17, 2019, this index employs a market capitalizationweighted methodology and comprises 502 USD-denominated constituent securities. Data prior to the launch date are based on back-tested calculations that replicate the index’s construction approach, though potential biases may exist due to historical influences on constituent selection. Nevertheless, these estimates provide a reasonable approximation of GF performance. Higher index values reflect more substantial progress in GF, while lower values suggest slower growth. To evaluate AI advancement, we use the S&P Kensho New Economy RACI, also obtained from S&P Global. This paper therefore Internal Auditing & Risk Management Year XX, No 2(72) September 2025 62 investigates the interrelationships among GF, AI, and ES, with their temporal trends visualized in Figure 1. 125 130 135 140 145 150 155 160 165 170 16 20 24 28 32 36 40 44 48 52 2016 2018 2020 2022 2024 GF (left) ES (right) 50 100 150 200 250 300 350 400 450 500 16 20 24 28 32 36 40 44 48 52 2016 2018 2020 2022 2024 AI (left) ES (right) 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 16 20 24 28 32 36 40 44 48 52 2016 2018 2020 2022 2024 WUI (left) ES (right) Figure 1. The trends of GF, AI, WUI, and ES Figure 1 reveals that the dynamics of the interactions between GF, AI, ES, and WUI possess complexities that may be difficult for standard analytical techniques to fully comprehend. Consequently, conventional approaches like SVAR and VAR are inadequate for capturing such dynamically evolving relationships, making the TVP-VAR-SV model better suited for quantitative analysis. Moreover, the strong correlation between WUI and factors like GF, AI, and ES means that any fluctuations in WUI significantly affect their interconnectedness. Therefore, we have incorporated WUI data as a control variable in our analysis. By scrutinizing the evolving impulse responses of GF, AI, and ES, while accounting for the influence of WUI, we can gain deeper insights into the time-varying relationships among GF, AI, and ES. Table 1. Descriptive statistics for GF, AI, WUI, and ES GF AI WUI ES Observations 120 120 120 120 Mean 142.565 219.586 24947.01 31.775 Median 140.362 196.063 22394.10 30.000 Maximum 160.954 413.990 57518.00 51.000 Minimum 127.130 100.011 9050.300 20.000 Standard Deviation 9.401 85.696 9540.084 7.119 Skewness 0.567 0.480 1.190 0.667 Kurtosis 2.078 2.266 4.469 2.579 Jarque-Bera 10.667 *** 7.298 ** 39.109 *** 9.797 *** Probability 0.005 0.026 0.000 0.007 Internal Auditing & Risk Management Year XX, No 2(72) September 2025 63 Table 1 presents the descriptive statistics of the variables, with mean values of 142.565, 219.586, 24947.01, and 31.775, indicating that the observations are generally clustered around these central tendencies. Nevertheless, the considerable spread between the maximum and minimum values reflects substantial fluctuations over time. The Jarque-Bera test results reject the normality assumption for GF, WUI, and ES at the 1% significance level, and for AI at the 5% level. To facilitate computation and reduce the influence of outliers, a natural logarithm transformation is applied. Furthermore, first-order differencing is employed for all variables to prevent spurious regression in the TVP-VAR-SV model. 5. Quantitative Discussions To ascertain whether GF, AI, ES, and WUI possess unit roots, we utilize three distinct tests: the ADF test, the PP test, and the KPSS test. Upon reviewing the results presented in the subsequent table, we can confidently prove that these variables do not exhibit unit roots. Table 2. Results of unit root tests ADF PP KPSS GF -4.559 (2) *** -7.243 [5] *** 0.135 [6] AI -7.284 (1) *** -8.672 [8] *** 0.137 [1] WUI -11.031 (1) *** -15.395 [8] *** 0.051 [9] ES -9.428 (2) *** -11.493 [4] *** 0.202 [10] Employing the SIC, a lag length of 2 was determined to be optimal for the TVP-VAR-SV model incorporating GF, AI, ES, and WUI. As shown in Table 3, the mean values of the estimates fall within the 95% confidence intervals. To verify estimation reliability, we ensured that the coefficients align closely with their posterior distributions. Moreover, inefficiency factors below 100 indicate that the system generates adequately independent samples. We therefore conclude that the parameter estimates are statistically credible. Table 3. The parameter estimated results in the TVP-VAR-SV process Parameters Mean Standard Deviation 95% Confidence Interval Geweke Inefficiency Factors 0.023 0.003 [0.018, 0.029] 0.239 5.44 0.022 0.002 [0.018, 0.027] 0.522 6.59 0.093 0.054 [0.043, 0.218] 0.327 82.98 Internal Auditing & Risk Management Year XX, No 2(72) September 2025 70 6. Conclusion Utilizing a methodology that incorporates the GF, AI, ES, and WUI, we can obtain estimated coefficient values, autocorrelation samples, trajectories, and posterior probability densities for the TVP-VAR-SV model. This technique enables us to accurately capture the evolving relations among the chosen indicators, thereby delving into the intricate dynamics between GF, AI, and ES. We find that in the initial stage, GF has a negative impact on ES, but in the second and third stages, it turns into a positive impact, showing the lag of its impact. The influence of AI on ES is negative in the early stage and the second stage, and it turns positive until the third stage, and the lag of its influence is more obvious. GF has shown a continuous positive promoting effect on AI. However, the influence of AI on GF is positive in the first phase, and showed a complex situation of interweaving positive and negative influences after entering the second and third phases. In addition, WUI has both positive and negative effects on ES, which verifies its rationality as a control variable to a certain extent. Through a detailed analysis of the dynamic relationships among GF, AI and ES, this study concludes that both GF and AI contribute positively to ES, although their influences exhibit inherent time lags. Following the aforementioned conclusions, a set of policy recommendations are put forth to safeguard energy security. Firstly, the government ought to augment its policy support and financial allocations for green finance, encompassing but not limited to the provision of incentives such as tax reliefs, subsidy schemes, and the establishment of dedicated funds. These measures are aimed at encouraging a greater influx of social capital towards green energy initiatives. This strategy will not only expedite the development and deployment of green energy technologies but also foster the construction of clean energy infrastructure, thereby laying a solid foundation for sustainable development. Meanwhile, it is necessary to establish and improve the green financial supervisory mechanism. Based on the formulation of environmental impact assessment standards and disclosure methods for publicly offering information, we can ensure that every green investment yields a win-win result in environmental and economic benefits while ensuring appropriate fund use and smooth project operation. Another key development toward a more energy-efficient future is the widespread dissemination of AI technology in the energy sector. For instance, smart grids using AI technology can adopt automatic grid scheduling, predict faults, and quickly respond to improve efficiency and stability in the energy system. In addition, AI can guarantee optimal distribution, reduce energy wastage, and balance supply and demand. In this regard, the government should further escalate support for research, development, and innovation in Internal Auditing & Risk Management Year XX, No 2(72) September 2025 71 AI technology, develop industry-university-research cooperation, cultivate professionals, and actively promote the application of new AI technologies in energy security continuously. This includes the forecast of energy demand by big data analysis and the optimization of energy production strategies through machine learning. Finally, the government should build a comprehensive policy framework for energy that could combine green finance with artificial intelligence development to capture synergies: setting up precise targets for energy transition, developing practical implementation paths and roadmaps, setting up crossdepartmental coordination mechanisms to ensure a coherent and efficient implementation of varied policies. The policy framework should also consider the balance between traditional and renewable energy development, ensuring the stability and security of energy supply. Strengthening international cooperation is equally important to achieve the long-term goal of energy security. In the face of global energy challenges, countries should increase experience sharing, technology exchanges, and market openness, promoting the global energy transition, addressing climate change, and achieving sustainable development goals together. This includes participation in international energy organizations, the signing of agreements on climate change, and facilitating cross-border collaboration on green energy projects to jointly build a greener, smarter, and safer global energy system. References Abadi, M.M.K.F., Liu, C., Zhang, M., Hu, Y., and Xu, Y. (2025). Leveraging AI for energy-efficient manufacturing systems: Review and future prospectives. J. Manuf. Syst. 78, 153-177. Arowoiya, V.A., Moehler, R.C., and Fang, Y. (2024). Digital twin technology for thermal comfort and energy efficiency in buildings: A state-of-the-art and future directions. Energy Built Environ, 5 (5), 641-656. Baseer, M.A., Kumar, P., and Nascimento, E.G.S. (2025). Advancements in hydrogen production through the integration of renewable energy sources with AI techniques: A comprehensive literature review. Appl. Energ. 383, 125354. Bashir, M.F., Pata, U.K., and Shahzad, L. ( 2025). Linking climate change, energy transition and renewable energy investments to combat energy security risks: Evidence from top energy consuming economies. Energy 314, 134175. Bhattacharyya, R. (2023). Enabling the clean energy transition: An assessment of the role of sustainable and green finance taxonomies. Reference Module in Social Sciences, Elsevier. Canova, F., and Gambetti, L. (2009). Structural changes in the US economy: Is there a role for monetary policy? J. Econ. Dyn. Control 33 (2), 477-490. Internal Auditing & Risk Management Year XX, No 2(72) September 2025 72 Chi-Chuan Lee, and Yuzhu Fang (2025). Climate finance for energy security: An empirical analysis from a global perspective. Economic Analysis and Policy,Volume 85, pp. 963-978. Choi, J.Y., Yim, H., Chi, S.Y., and Lee, M.J. (2024). Future scenarios of digital technology-driven energy consumption in South Korea. Energy Rep. 11, 908913. Choi, S.Y., and Hadad, E. (2025). The dynamic relationship among economic and monetary policy, geopolitical risk, sentiment, and risk aversion: A TVP-VAR approach. Financ. Res. Lett. 72, 106532. Dong, X.T., Wong, X.Q., Qin, M., Lobont, O.R., and Umar, M. (2025). Forest investment in China: an efficient way to climate and economic security? Econ. Change Restruct. 58 (1) , 7. Ghenai, C., Husein, L.A., Nahlawi, M.A., Hamid, A.K., and Bettayeb, M. (2022). Recent trends of digital twin technologies in the energy sector: A comprehensive review. Sustain. Energy Technol. Assess. 54, 102837. Ha, L.T., Bouteska, A., Sharif, T., and Abedin, M.Z. (2024). Dynamic interlinkages between carbon risk and volatility of green and renewable energy: A TVP-VAR analysis. Res. Int. Bus. Financ. 69, 102278. Huang, C., and Lin, B. (2023). Promoting decarbonization in the power sector: How important is digital transformation? Energ. Policy 182, 113735. Iqbal, M.A., Saheen, W.A., Shabir, S., Ullah, U., Ionel-Alin, I., Mihut, M.I., Raposo, A., and Han, H. (2025). Towards a green economy: Investigating the impact of sustainable finance, green technologies, and environmental policies on environmental degradation. J. Environ. Manage. 374, 124047. Khan, K. (2025). How do supply chain and geopolitical risks threaten energy security? A time and frequency analysis. Energy 316, 134501. Kim, J., Jaumotte, F., Panton, A.J., and Schwerhoff, G. (2025). Energy security and the green transition. Energ. Policy 198, 114409. Kuang, M., Kuang, D., Rasool, Z., Saleem, H.M.N.,and Ullah, M.I. (2024). From bytes to sustainability: Asymmetric nexus between industrial artificial intelligence and green finance in advanced industrial AI nations. Borsa Istanb. Rev. 24 (5), 886-897. Liu, Y., Dong, K., and Nepal, R. (2025). Can sustainable financing facilitate the energy justice transformation? Evidence from developing countries in Asia. Glob. Financ. J. 101081. Liu, Y., Zhao, X.M., Dong, K.Y., and Jiang, Q.Z. (2023). Assessing the role of green finance in sustainable energy investments by power utilities: Evidence from China. Util. Pol. 84, 101627. Lu, J., and Li, H. (2023). Can digital technology innovation promote total factor energy efficiency? Firm-level evidence from China. Energy 293, 130682. Madaleno, M., Dogan, E., and Taskin, D. (2022). A step forward on sustainability: The nexus of environmental responsibility, green technology, clean energy and green finance. Energ. Econ. 109, 105945. Internal Auditing & Risk Management Year XX, No 2(72) September 2025 73 Maghyereh, A., Ziadat, S.A., and Rababa’a, A.R.A. (2024). Exploring the dynamic connections between oil price shocks and bond yields in developed nations: A TVP-SVAR-SV approach. Energy 306, 132475. Mohammadi, M., and Sohn, I. 2023. AI based energy harvesting security methods: A survey. ICT Express 9 (6), 1198-1208. Nakajima, J., Kasuya, M., and Watanabe, T. (2011). Bayesian analysis of time-varying parameter vector autoregressive model for the Japanese economy and monetary policy. J. JPN. Int. Econ. 25 (3), 225-245. Nepal, R., Zhao, X., Dong, K., Wang, J., and Sharif, A. (2025). Can artificial intelligence technology innovation boost energy resilience? The role of green finance. Energ. Econ. 142, 108159. Nepal, R., Zhao, X., Liu, Y., and Dong, K. (2024). Can green finance strengthen energy resilience? The case of China. Technol. Forecast. Soc. Change 202, 123302. Primiceri, G.E., 2005. Time varying structural vector autoregressions and monetary policy. Rev. Econ. Stud. 72 (3), 821-852. Qiao, S., Dang, Y.J., Ren, Z.Y., and Zhang, K.Q. (2023). The dynamic spillovers among carbon, fossil energy and electricity markets based on a TVP-VAR-SV method. Energy 266, 126344. Qin, M., Hu, W., Qi, X.Z., and Chang, T.Y. (2024). Do the benefits outweigh the disadvantages? Exploring the role of artificial intelligence in renewable energy. Energ. Econ. 107403. Qin, M., Wu, T., Ma, X.C., Albu, L.L., and Umar, M. (2023). Are energy consumption and carbon emission caused by Bitcoin? A novel time-varying technique. Econ. Anal. Policy 80, 109-120. Rodriguez, G., Castillo, B.P., Calero, R., Cisneros, R.S., and Arellano, M.A. (2024). Evolution of the exchange rate pass-through into prices in Peru: An empirical application using TVP-VAR-SV models. J. Int. Money Financ. 142, 103023. Sevillano, M.C., Jareño, F., López, R., and Esparcia, C. (2024). Connectedness between oil price shocks and US sector returns: Evidence from TVP-VAR and wavelet decomposition. Energ. Econ. 131, 107398. Shi, Y., and Zhao, Y. (2023). The contribution of green finance to energy security in the construction of new energy system: Empirical research from China. J. Clean. Prod. 429, 139480. Shupler, M., Mwitari, J., Gohole, A., Cuevas, R.A., Puzzolo, E., Čukić, I., Nix, E., and Pope, D. (2021). COVID-19 impacts on household energy & food security in a Kenyan informal settlement: The need for integrated approaches to the SDGs. Renew. Sustain. Energy Rev. 144, 111018. Usman, A., Naqvi, S.M.M.A., Ozturk, I., Hassan, A., and Arif, A. (2025). ICT-driven urbanization and energy security risk: Empirical evidence from Group 7 and Emerging 7 economies. Environ. Impact Assess. Rev. 112, 107809. Wang, P., and Xu, X. (2025). Green finance and energy efficiency improvement: The role of green innovation and industrial upgrading. Innov. Green Dev. 4 (1), 100200. Internal Auditing & Risk Management Year XX, No 2(72) September 2025 74 Xiong, Y., and Dai, L. (2023). Does green finance investment impact on sustainable development: Role of technological innovation and renewable energy. Renew. Energy 214, 342-349. Yu, W., Patros, P., Young, B., Klinac, E., and Walmsley, T.G. (2022). Energy digital twin technology for industrial energy management: Classification, challenges and future. Renew. Sustain. Energy Rev. 161, 112407. Zeng, L., Wong, W.K., Fu, H., Mahmoud, H.A., Cong, P.T., Thuy, D.T.T., and Bach, P.X. (2024). FinTech and sustainable financing for low carbon energy transitions: A biodiversity and natural resource perspective in BRICS economies. Resour. Policy 88, 104486. Zhai, M., Wu, W., and Tsai, S.B. (2025). The effects of Artificial intelligence orientation on inefficient investment: Firm-level evidence from China’s energy enterprises. Energ. Econ. 141, 108048. Zhang, C., Wang, C., Jin, H., Xu, M., Balezentis, and T., Agnusdei, L. (2025). The impact of green finance on energy transition and carbon emission targets: Exploring the double threshold and spatial spillover effects among the regions. Struct. Change Econ. Dyn. 73, 1-10. Zhang, J., and Sun, Z. (2025). Energy-environmental efficiency enhancement using green finance through spatio-temporal heterogeneity and dynamic regulatory mechanisms: Multiple perspectives study. Renew. Energ. 240, 122212. Zhang, L., Saydaliev, H.B., and Ma, X. (2022). Does green finance investment and technological innovation improve renewable energy efficiency and sustainable development goals. Renew. Energy 193, 991-1000. Zhao, Z., Zhao, Q., Li, S., and Yan, J. (2025). Artificial Intelligence-driven regional energy transition:Evidence from China. Econ. Anal. Policy 85, 48-60. Zhong, Y.F., Chen, X.F., Wang, C.F., Wang, Z.X., and Zhang, Y.C. (2023). The hedging performance of green bond markets in China and the U.S.: Novel evidence from cryptocurrency uncertainty. Energ. Econ. 128, 107194. Zhou, Y., Li, X., Wu, Z., Wu, J., and Li, H. (2025). Green bonds and intelligent manufacturing: Evidence from listed firms in China. Econ. Lett. 247, 112150. Zhu, X., Zhang, S., and Ding, Q. (2024). Does extreme climate change drive the connectedness among global gold markets? Evidence from TVP-VAR and causality-in-quantiles techniques. Resour. Policy 91, 104899.