A novel hybrid MCDM approach for evaluating digital technologies for sustainability transition in the banking industry

Digital technologies are being implemented across value chains at an accelerating rate as the urgent clamor for digital disruption of bank ecosystems grows more robust (Sibanda, Ndiweni, Boulkeroua, Echchabi, & Ndlovu, 2020). The banking industry embraces the potential metamorphosis of disruptive technological advancements (Osei, Cherkasova, & Oware, 2023) through the confluence of multiple digital technologies, including artificial intelligence-driven machine learning, the Internet of Things, cloud banking, biometrics, and information technology (Mehdiabadi, Tabatabeinasab, Spulbar, Karbassi Yazdi, & Birau, 2020; Naimi-Sadigh, Asgari, & Rabiei, 2022). Remarkably, the proliferation of these technologies streamlines payment processes (Balkan, 2021), elevates operational efficacy (Choubey & Sharma, 2021), constrains illicit activities (Tsindeliani et al., 2022), and facilitates fraud detection while also dealing with customer relationship management (Boateng, Adam, Okoe, & Anning-Dorson, 2016). In parallel, banks are facing mounting pressure to incorporate sustainability considerations (Bukhari, Hashim, & Amran, 2022) in response to global economic expansion, deepening social inequalities, and diminishing natural resources (Bocken, Short, Rana, & Evans, 2013). These demands may stem internally from stakeholders (Kumar & Prakash, 2020) or externally, from regulatory bodies (Dubey et al., 2019) and customers (Ahuja, 2015). Given the escalating challenges, numerous banks and financial institutions have adjusted their business models to integrate a triple bottom line (TBL) sustainability model, encompassing environmental, social, and economic aspects. Where the environmental metric assesses the influence of the bank's operational activities on the environment, the inculcation of social indicators evaluates the impact on social concerns, and lastly, the economic parameter quantifies the bank's profitability and long-term viability (Azouaoui, Berjaoui, & Houssaini, 2023). Also, banks have reimagined business strategies that harness digital technologies to uphold their sustainability and the sustainability of an exemplary financial system (Çokçetin, 2017).

Conceptually, digital technologies allude to a paradigmatic convergence of multiple disruptive and groundbreaking technologies in the era of Industry 4.0 (Li, Dai, & Cui, 2020). The evolution of technology in the 4.0 era has provided an irrepressible impetus for banking systems to diversify their offerings by integrating online banking services (Garg & Kumar, 2024). In this dynamic ecosystem, banks cannot thrive and persist if they are cognizant of how their actions influence the sustainability of their business (Ameen, 2017). Accordingly, the rolling out of contemporary technologies in the financial industry enables the production of additional value through enhanced offerings, innovative business models, and augmented operational efficiency, thereby opening the door to sustainable development (Mavlutova et al., 2022). Over the past decade, commercial banks' average total factor productivity surged by 10.7% due to technical efficiency and advancement (Zuo, Strauss, & Zuo, 2021). Concerning this, authors Kolodiziev, Shcherbak, Chernovol, and Lozynska (2022) confirmed that the developed approach for digitalizing financial services can elevate customer loyalty by 15%, reduce risk by 10%, and make banks more intriguing to investors by 15–20%. Besides, blockchain is transforming the banking industry by managing massive volumes of data. Large banking and financial institutions are implementing this technology to combat financial fraud due to its traceability, immutability, and decentralization attributes (Mishra & Kaushik, 2023). Moreover, authors Hopalı, Vayvay, Kalender, Turhan, and Aysuna (2022) considered mobile wallets crucial for sustainability since traditional checkout payments emit an average of 3.78 g of carbon dioxide in every transaction. Likewise, several studies pointed out that robotics is changing the future of banking and promoting sustainable practices (Ciufudean, 2018; Choubey & Sharma, 2021). From this point forward, we can affirm that digitalization, digital technology, and digital transformation hold significant promise for enhancing the TBL sustainability of the banking sector (Broccardo, Zicari, Jabeen, & Bhatti, 2023; Garg & Kumar, 2024).

Now, the question of which specific digital technologies are making the banks more sustainable surfaces. Concerning this, prior literature in their studies have applied the analytical hierarchy process (AHP) model for evaluating the social sustainability of the technology management process (Nejad, Mansour, & Karamipour, 2021), recognized impediments to industry 4.0 sustainable digital technologies (Verma, Kumar, Daim, Sharma, & Mittal, 2022), identified connected vehicle technologies for sustainable development (Nasrollahi, Ghadikolaei, Ghasemi, Sheykhizadeh, & Abdi, 2022), prioritized blockchain adoption factors by incorporating fuzzy AHP with TOE framework (Kajla, Sood, Gupta, Raj, & Singh, 2023), determined organization agile capabilities for sustainable digital transformation (Feroz, Zo, Eom, & Chiravuri, 2023) and emphasized the implications of Industry 4.0 on the attainment of sustainable development goals in manufacturing operation (Agarwal & Ojha, 2024). Upon review, the majority of academic research production was narrowly focused and accentuating particular concepts (like Industry 4.0), technologies (like blockchain technology or artificial intelligence), industries (like manufacturing), and concentrating on specific aspects of TBL sustainability. To the best of the author's knowledge, no study explicitly identifies digital technologies contributing to the sustainability shift, particularly in the banking sector, through a systematic literature review (SLR) and prioritizes these sustainable digital technologies by utilizing the multi-criteria decision-making (MCDM) approach, specifically the fuzzy Delphi method (FDM), and best-worst method (BWM). Unquestionably, digital technologies are one of the fastest-growing phenomena, but their impact on banking sustainability is still not widely recognized (George, Merrill, & Schillebeeckx, 2021; Mavlutova et al., 2022; Broccardo et al., 2023; Garg & Kumar, 2025). There is a dearth of studies addressing all three aspects simultaneously.

The hybrid integration of the SLR-FDM-BWM moves the study from a comprehensive and unbiased approach to expert-based validation to precise prioritization that provides robustness, minimizes bias, and improves decision accuracy. SLR confirms that the identified sustainable digital technologies criteria are based on evidence-based screening rather than subjective selection. Then, FDM filters the technologies by using fuzzy logic to handle inconsistency and expert consensus-building. The validated sustainable technologies are finally given strong weights by the BWM with less comparison burden in comparison to AHP. This multi-level approach provides a replicable hybrid methodology that other technical sectors can adopt for targeting sustainability objectives. The study is novel both in methodological integration and implementation in a cross-disciplinary setting.

In a nutshell, we envisioned investigating the following study questions:

A meticulous and exhaustive assessment of the literature confers upon the authors an edge over other scholarly investigations through:

1. The postulated novel methodology leverages a Delphi method in conjunction with fuzzy set theory to circumvent the uncertainties in filtering out digital technologies. The FDM applied in this work is single-round; a consensus can be reached without the need for further rounds;

2. Further, BWM has been applied to weigh the different sustainable digital technologies. In this regard, Rezaei (2015) asserts that in comparison with the widely used weighting approach, namely AHP, it has several advantages since it yields more consistent findings with fewer expert judgments and comparisons involved;

3. Attempting to deliver answers to research queries derived from currently published studies on these concerns;

4. Contributing to the existing corpus of research literature focused on integrating both concepts.

The remaining portion has been designed as follows. Section 2 comprises an exhaustive review of research about digital technologies propelling sustainability in the banking sector. Critical sustainable digital technologies have been pinpointed in this section. Section 3 highlights the research methodology and defines the proposed research framework. Section 4 provides an expanded analysis of the FDM-BWM results. Section 5 briefly discusses the findings, highlighting the research implications and laying the groundwork for final observations, along with major limitations and future avenues for exploration in Section 6.

We have performed an SLR to determine the prominent dig-techs driving the banking industry's shift to sustainability. Specifically, SLR differs from other literature review methodologies primarily due to its openness, inclusivity, explanatory nature, and heuristic nature. These characteristics enable objective evaluation of search results and the elimination of any prejudices and inconsistencies present within the data (Denyer & Tranfield, 2009; Garg & Kumar, 2024). In this regard, the authors followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework by Moher (2009) to evaluate the data and streamline it meticulously (Garg & Kumar, 2025) (refer to Figure 1). The primary objective of PRISMA is to present literature reviews clearly and transparently. A review protocol was designed under these guidelines, addressing the research strategy for searching the database, article inclusion criteria, and quality evaluation.

The authors have preliminarily examined relevant literature through the Scopus database (Link to the website, accessed on 11th April 2024) and have developed strings encapsulating keywords to satisfy our RQs. Papers were extracted from the database using various keyword combinations related to digital technologies, digitalization, sustainability, the triple bottom line, and banking, with the search limited to “titles, abstracts, and/or keywords”. Consequently, this search produced 283 matches in the Scopus database. The inclusion criteria listed in Table 1 were incorporated to minimize the likelihood of including false positives (Garg & Kumar, 2023). In this process, 164 articles meeting the exclusion criteria were discarded. Subsequently, 119 articles were eligible for abstract review evaluation. After examining the abstracts, we determined 80 papers for full-text review assessment.

An extensive examination of the entire texts of the shortlisted studies was conducted by two experts to assess their quality and mitigate potential repercussions. The first expert is an academician with ten years of experience in banking and financial technologies, and the second expert is a bank manager with twelve years of experience in the banking industry, possessing concrete knowledge of operational assessment and technology adoption. For this, the quality assessment criteria listed in Table 2 were employed to analyze the relevance, veracity, and completeness of 80 documents. However, these criteria have been adapted from the study of Nguyen-Duc, Cruzes, and Conradi (2015). Each criterion encompasses four possible scores: 0 for no mention, 1 for restricted mention, 2 for adequately addressed, and 3 for comprehensive mention. Only research with an average quality score of at least one was considered to ascertain the validity of selected articles.

Based on this, the authors narrowed down the final dataset of 64 articles to determine the crucial technologies enabling sustainability in the banking industry. In addition, these dig-techs were meticulously discussed with academicians, professionals having sufficient proficiency in the field of the current study, and chief bank managers (IT and operations) with deep expertise in digital technologies. After performing SLR and in detailed discussions with experts, nine digital technologies propelling banking sustainability have been proposed, details of which are presented in Table 3. An illustrative depiction of the critical sustainable digital technologies is illustrated in Figure 2. Below is a summary of the identified digital technologies.

Artificial intelligence (AI) has been described as an integrated mental capacity for acquiring knowledge, thinking, and problem-solving according to Snyderman and Rothman (1987). AI-based workflow procedures have the potential to supplant ineffective paperwork workflows, offering more effective ways to store, distribute, manage, and exploit data while drastically reducing expenses and processing time in the banking sector (Hoang, Nguyen, & Le, 2022). The application of AI in banking can enhance infrastructure project risk assessment models, supporting funding for sustainable development projects (Elias et al., 2024). It is perceived as the way of the future since it strengthens compliance and delivers innovative financial solutions for halting illicit financial activities (Rabbani, Lutfi, Ashraf, Nawaz, & Ahmad Watto, 2023). However, prior studies outlined AI and its connotation with the particular aim of sustainable development goals and encountered an affinity between AI and the UN 2030 Agenda for sustainable development (Di Vaio, Palladino, Hassan, & Escobar, 2020; Chadha & Mehta, 2022; Elias et al., 2024).

According to the World Economic Forum, Blockchain technology (BC) is considered to be the principal driving force behind the financial industry (Mishra & Kaushik, 2023). This technology has been designed to store data digitally by leveraging a peer-to-peer network architecture (Rejeesh & Thejaswini, 2020). The fundamental objective behind this technology is to create an accessible, transparent, open, and decentralized ledger that is accessible to all users, eradicating the need for intermediaries (Schuetz & Venkatesh, 2020; Mehdiabadi, Tabatabeinasab, Spulbar, Karbassi Yazdi, & Birau, 2020; Hoang et al., 2022; Mbaidin, Alsmairat, & Al-Adaileh, 2023; Mishra & Kaushik, 2023; Gupta, 2023). The implementation of this technology implies the security of every transaction and dampens the likelihood of possible fraud committed by altering the distinctive identity (Giungato, Rana, Tarabella, & Tricase, 2017). Besides, BC elevates performance and competitiveness as part of an ambidextrous strategy, mitigates operational risks through interbank reconciliation, and emphasizes information processing (Azouaoui et al., 2023). Further, Wang, Zhao, and Chen (2023) study posited that the application of BC might augment the global financial infrastructure and promote sustainable growth by installing more efficient mechanisms.

Big data analytics (BDA) is the development of novel architectures and technologies intended to systematically extract value from massive amounts of data by encapsulating high velocity, exploration, and/or analytics (Mikalef, Pappas, Krogstie, & Giannakos, 2018; Ali, Salman, Yaacob, Zaini, & Abdullah, 2020). The banking industry benefits greatly by exploiting the power of big data, as it enables faster response to consumer needs and promotes engagement throughout the whole process. This reduces expenses, CO2 emissions, and energy consumption, making the banking activities more sustainable (Hahn, Pinkse, Preuss, & Figge, 2015; Azouaoui et al., 2023). According to Zhu and Yang (2021), Asian banks stimulate banks to take ethical and social concerns into their operations with the help of BDAs to develop sustainable finance policies, guidelines, and strategies.

Application programming interfaces (API) are collections of rules for programs that facilitate communication with each other and serve as a conduit between various applications (Mehdiabadi et al., 2020; Gupta, 2023). Implementing APIs permits banks to work with third parties and other stakeholders, such as fintechs, to accomplish their digital goals and develop revolutionary products (Gupta, 2023). Besides, open banking based on API technology gives the banks the liberty to set parameters for the amount of shared data (Billiam, Abubakar, & Handayani, 2022). API-enabled open banking also reduces the environmental impact of conventional in-branch banking operations by enhancing user experience and promoting ongoing usage of digital channels (Althinayyan & Alojail, 2024).

The emerging financial transactions enabled by Mobile banking (MB), commonly known as mobile transfers, m-banking, or mobile money, are increasingly being assimilated into the global monetary framework (Hoang et al., 2022). A significant number of banks increasingly provide MB based applications, which ultimately raises market share and profitability and offers economic advantages (Mullan, Bradley, & Loane, 2017). In light of the quantity of carbon dioxide released by customary checkout payment procedures, mobile wallets provide a secure, dependable, and trackable payment solution essential for sustainability. A study by Hopalı et al. (2022) disclosed that mobile wallets enable usefulness, security, cost-effective services, transaction speed, and a cashless transactional approach that optimizes payment service sustainability.

Robotics process automation (RPA) is a software solution that makes use of software bots to automate rule-based business procedures (Lacity, Willcocks, & Craig, 2015; Kokina & Blanchette, 2019; Kregel, Koch, & Plattfaut, 2021). They can lessen functional mistakes, operate continuously around the clock, strengthen compliance, and be better positioned for executing repetitive operations (Vijai, Suriyalakshmi, & Elayaraja, 2020; Mehdiabadi et al., 2020). This augments sustainability and spurs growth and innovation in consumer services (Choubey & Sharma, 2021). However, robotics has the best prospects for banking sustainability holistically since it requires no additional infrastructure and uses a low-code methodology.

The Internet of Things (IoT) can be defined as a worldwide network of connected items that can only be addressed using standard communication protocols (Beier, Niehoff, & Xue, 2018). IoT has many devourers in the financial services industry, including smart customer navigation, prompt interaction with customers, upgraded physical security measures, and increased corporate efficiency (Hoang et al., 2022). Accordingly, banks should consider embracing IoT technologies to slash fixed and operating expenses and fortify their position in the market (Mital, Chang, Choudhary, Papa, & Pani, 2018). The IoT ecosystems present viable ways to save energy and communications costs while maintaining data privacy, highlighting the promising application of environmentally friendly IoT designs in the banking industry (Qi & Hossain, 2024).

Regulatory Technology (RegTech) integrates digital and information technologies to improve regulatory processes, including reporting, compliance, and regulatory monitoring (Becker, Merz, & Buchkremer, 2020; Von Solms, 2021). The process can become more resilient and cost-effective by standardizing, automating, and expediting numerous manual tasks through RegTech (Von Solms, 2021). With this, banking institutions can reduce the expenses associated with regulatory compliance and can better manage risks (Quill & Lennon, 2019). Moreover, the banking sector can reinforce the economic, social, and environmental facets of sustainability by strengthening disclosure procedures, lowering compliance costs, and enhancing internal controls (Kanojia, Kaur, & Bhavya, 2024).

The establishment of on-demand access to a pool of programmable computer resources comprising servers, networks, storage, apps, and services is referred to as Cloud computing (CC) (Łasak & Wyciślak, 2023). It delivers an inexhaustible infrastructure for data storage and execution and streamlines customer relationship management (Azouaoui et al., 2023; Gupta, 2023). The user can rent a portion of the public cloud or have their private cloud (Gupta, 2023). As an internet-based technological resource paradigm, CC provides specific accessibility, real-time data, and information transfer, expedited deployment, higher adaptability, and storage capabilities (Yoon, 2020). Consequently, it offers value for both external sustainability reporting and internal sustainability accounting (Vărzaru, 2022).

Research on sustainable digital technologies, particularly within the banking sector, is still in its infancy (George et al., 2021; Broccardo et al., 2023; Garg & Kumar, 2025). The majority of academic research on this topic has been narrowly focused, accentuating particular concepts (such as Industry 4.0), technologies (like blockchain technology or artificial intelligence), and industries (like manufacturing), while concentrating on specific aspects of TBL sustainability. Although few studies in the past have explicitly highlighted the positive interrelationship (Mavlutova et al., 2022; Mir & Bhat, 2022; Dunbar, Sarkis, & Treku, 2024; Garg & Kumar, 2024, 2025; Hu, Chun-Wei, & Hariyadi, 2024) between digital technologies and triple bottom line sustainability within the Indian banking sector. However, there is a notable lack of studies in the banking sector domain that systematically identify key sustainable technologies from the literature using the SLR process, validate them through FDM, and weigh them through BWM. To fill these research gaps, the proposed research aims to establish a benchmark that may be extensively applied and expanded for further investigation.

Table 4 provides a summary of recent applications of the FDM and BWM across digital technologies and sustainability sectors to clearly establish the methodological contribution of this study. This comparison indicates their complementary capabilities and the justification for combining the two approaches in the current study.

Table 4 highlights that FDM is primarily used for expert consensus building and indicator screening, whereas BWM is used to determine priority weights for confirmed indicators. These approaches have only been applied successively in a few numbers of studies, especially when it comes to Banking 4.0 and triple bottom line sustainability. To close this methodological gap, a sequential hybrid FDM and BWM is employed.

A novel three-phase methodology is put forward to unleash digital technologies driving sustainability in the banking industry, as portrayed in Figure 3. First, a systematic literature study was conducted to extract pertinent documents from the Scopus database that synthesized the keywords associated with the digital technologies employed by banks and their impact on the sustainability framework. Subsequently, following deliberations with an expert panel of six bank managers and three academicians, nine digital technologies were refined with an eye toward their applicability and practicality, as indicated in Table 3. Demographic characteristics of experts are given in Table 5. Second, the FDM was applied to assess the significance of each digital technology in the sustainability shift based on the Indian instance. The FDM technique presented here was devised to critically assess a list of proposed sustainable digital technologies (see Table 3) extracted from the SLR and the expert panel's feedback. The selection criteria of experts met the recommendations of Adler and Ziglio (1996), which included sufficient time to participate, being willing and able to participate, having credentials and an understanding of the topics being investigated, and having strong communication skills. Third, the weights and preferences of each sustainable digital technology were determined through the execution of the BWM, a relatively new technique put out by Rezaei (2015) that addresses the shortcomings of the widely used analytical hierarchy process (AHP) weighing method. This approach is acknowledged for its efficacy in studies involving a limited number of specialists (Rezaei, 2015).

FDM is a formal communication methodology or strategy that was originally proposed to be an interactive, systematic, and extrapolative process that relied upon the opinions of an expert panel (Dalkey & Helmer, 1963; Hsu & Sandford, 2007). The FDM is a method for gathering expert opinions that consists of three primary components: anonymous response, controlled feedback iteration, and statistical group response (Hsu, Lee, & Kreng, 2010). However, expert opinions cannot be reliably converted into numerical measurements. Therefore, crisp values cannot accurately represent practical systems due to the equivocal, subjective nature of human thought, preferences, and judgments (Kannan, de Sousa Jabbour, & Jabbour, 2014). Consequently, to address the ambiguity of human cognition and behavior in decision-making, Zadeh (1978) invented the fuzzy set theory (Shen, Olfat, Govindan, Khodaverdi, & Diabat, 2013). This FDM synthesizes the traditional Delphi technique with fuzzy set theory to address some of the Delphi Consensus panel's uncertainty and enhance the Delphi method's inadequacy and vagueness. The FDM has been carried out by executing the following steps:

• Step 1: Determination of digital technologies supporting sustainability in banking. Initially, a systematic assessment of the literature determined which digital technologies might cause a shift in banking towards sustainability.

• Step 2: Identification of expert perspectives with a decision-making group. Following the recognition of digital technologies, several industrial and academic decision-makers were approached as experts to assess the severity of digital technologies using a questionnaire and the linguistic variables specified in Table 6. Triangular fuzzy numbers (TFN) of a five-point Likert scale ranging from “very unimportant” to “very important,” given by Habibi, Jahantigh, and Sarafrazi (2015), were used in this study to evaluate digital technologies (see Figure 4). Furthermore, the expert group judgment was ascertained using a geometric mean approach (Ma, Shao, Ma, & Ye, 2011).

• Step 3: The recognition of significant digital technologies. The final stage of FDM sought to examine the important digital technologies by comparing the weight of each digital technology with the threshold value. $α$⁠. Each digital technology TFN $τ~j$ It was ascertained as follows:

In the preceding equations, the expert is represented by index i while the criterion is defined by index j. The fuzzy value of each criterion that each expert provided is symbolized by the notation $a~ij$⁠. The value of $τ~j$ Represents the fuzzy average value for each criterion. Furthermore, each criterion fuzzy average value is defuzzified into a crisp value (Habibi et al., 2015), which is equivalent to:

Notably, previous research has assigned varying numerical values to the threshold value. $α$⁠. In this regard, Petrudi, Tavana, and Abdi (2020) argued for four approaches for determining the threshold values: strict, mean, moderate, and conservative. The experts reached a consensus within this study to establish a threshold value of 5 or 0.5 in percentage terms. Accordingly, we are utilizing the moderate approach to ascertain that an appropriate number of digital technologies are chosen for additional analysis. Following the computation of the value as mentioned above, criterion j is accepted and moved on to the subsequent research phase if the crisp value of $τ~j$ is above the threshold $α$⁠. The criterion j is rejected if the crisp value of $τ~j$ is smaller than the threshold $α$⁠.

BWM is a comparative-oriented MCDM approach that assesses each criterion against the best criterion and all other criteria against the worst criterion. The comparison system constructs a basic linear optimization model to determine the ideal weights and consistency ratio (Rezaei, Nispeling, Sarkis, & Tavasszy, 2016; Ghaffari, Arab, Nafari, & Manteghi, 2017). This strategy comprises weighing criteria through pairwise comparisons similar to the AHP and analytic network process (ANP) (Saaty, 2004). Compared to AHP and ANP, this technique uses fewer pairwise comparisons and smaller sample sizes. Furthermore, the application of BWM outperforms other MCDM approaches, namely DEMATEL, AHP, TOPSIS, VIKOR, and ELECTRE. However, BWM results are more reliable and consistent than other MCDM techniques due to their simple and standardized approach to data collection for pairwise comparisons. Such data can be altered to enhance consistency (Kaushik, Kumar, Gupta, & Dixit, 2022; Kapoor, Sindwani, & Goel, 2022). Many scholars have applied BWM to prioritize online apparel return factors (Kaushik et al., 2022), determine the relevance of digital supply chain on sustainability (Arman & Organ, 2021), assess mobile banking applications (Roy & Shaw, 2021), evaluate blockchain technology adoption in the oil and gas industry (Munim, Balasubramaniyan, Kouhizadeh, & Hossain, 2022), uncover supply chain disruptions during COVID-19 (Ali, Charles, Modibbo, Gherman, & Gupta, 2023) and explore transformative strategies for cleaner production in industry 5.0 era (Sharma & Gupta, 2024). The steps of the BWM applied for the present study are elucidated below (Rezaei, 2015; Kaushik et al., 2022; Kapoor et al., 2022; Rajeb et al., 2022).

• Step 1: Develop a set of evaluation criteria. This step involves establishing evaluation criteria {C1, C2, . . ., Cn} for decision-making. A set of n criteria is identified after an exhaustive literature review.

• Step 2: Determine the best and worst criteria. In this stage, experts identify the best and worst criteria from the set of criteria finalized in the previous step. According to experts, the most influential criterion is considered the best, and the least influential is regarded as the worst.

• Step 3: Examine the preference for the best criterion over the other criteria. In this step, experts are asked to rate the best criterion on a scale of 1–9 identified in the prior step. Where 1 is “equally important” and 9 is “extremely more important”. The resulting best to other vectors would be $AB={aB1,aB2,aB3,......aBn},$ where $aBj$ suggests the preference of criteria B (best criteria) over criteria j and $aBB=1$⁠.

• Step 4: Evaluate the preference of the criteria over the worst criterion. This case is assigned a numerical value between 1 and 9 by experts. The resulting others to the worst vector would be $AW={a1W,a2W,a3W,......,anW}T$⁠, where $ajW$ implies a preference for criterion j over the worst criterion W and $aWW=1$⁠.

• Step 5: Compute the optimal weights $(w1*,w2*,w3*,......,wn*)$⁠. To get optimum weights, maximum absolute differences for all $j$ could be minimized for ${|wB−aBjwj|,|wj−ajWwW|}$⁠. Given below minimax model will be derived:

min max ${|wB−aBjwj|,|wj−ajWwW|}$

Model (1) has been transformed into a linear model to achieve better outcomes. The model has been presented below:

For obtaining optimal weights $(w1*,w2*,w3*,......,wn*)$ and optimal value $ξL$⁠, Model (2) could be solved. Consistency (⁠$ξL$⁠) of criteria comparisons near 0 is preferred (Razaei, 2015).

As explained previously in the methodology section, the FDM technique has been employed to determine the most significant digital technologies from the literature. For this purpose, a questionnaire was developed based on content analysis using a fuzzy triangular mean, and the amount of fuzzy decoding was determined in each stage. The results of FDM are documented in Table 7. We can observe from the table that the defuzzified value of the Application Programming Interface falls below the threshold value (⁠$τ~j$ < 0.5). Consequently, this technology has been excluded from further analysis. Additionally, Table 8 summarizes the digital technologies that underpin sustainability in the banking sector, drawing from FDM results and expert perspectives.

After finalizing the technologies, the importance weight was evaluated using BWM. Nine experts provided data in the form of opinions (six bank managers and three academicians). Following the BWM process outlined in the methodology, experts were requested to recognize the best and worst technology among the indicators (see Table 8) on a scale of 1–9. The comprehension of the scale has been addressed in Table 9 (Gupta, Anand, & Gupta, 2017). Subsequently, the decision panel was asked to prioritize the best criterion among other criteria and determine the preference of other criteria over the worst criterion. The experts' inputs yielded best-to-others and others-to-worst vectors, as apparent in Tables 10 and 11, respectively. The final weights of the criterion were derived using BWM's linear model after getting all the ratings from experts. We calculated the BWM linear model to determine the optimal weights and consistency value (⁠$ξL$⁠).

The final weight of the criteria was obtained by taking the arithmetic mean of the weights assigned to each expert, as evident in Table 12. The application of BWM unleashes AI (0.24577), MB (0.21517), BC (0.19274), CC (0.09554), RPA (0.0948), and BDA (0.05979) as the most important technologies, while IoT (0.05221) and RegTech (0.04396) are considered the least important, propelling sustainability in the banking industry. The results suggest that although the industry places heavy concentration on intelligence-driven and customer-facing solutions, the application of Internet of Things and regulatory technology capabilities remains confined and underutilized. This may be due to the early stages of extensive IoT integration in banking or possible issues related to interoperability and data protection. The intricacy of the regulatory environment and sluggish adoption of innovative regulatory technologies may contribute to the minimal impact that RegTech has on sustainable banking practices despite offering solutions for compliance and regulatory procedures. Further, the average consistency value (⁠$ξL$⁠) was discovered to be 0.0596, which is nearly zero. Hence, the comparisons are reliable and highly consistent.

The application of the multi-step methodology SLR-FDM-BWM has provided significantly deeper insights into the prioritization of different technologies impacting sustainability within the banking industry. The weighted assessment of digital technologies unveils a distinct hierarchy of technological relevance, including AI (0.24577), MB (0.21517), BC (0.19274), CC (0.09554), RPA (0.09480), BDA (0.05979), IoT (0.05221), and RegTech (0.04396). This directly addresses the research gap noted in prior literature. The results of the study are somewhat consistent with those of Mavlutova et al. (2022), who identified AI, MB, BC, CC, and RPA as the critical sustainable digital technologies. However, they are in contrast with previous literature, including the work of Bai, Dallasega, Orzes, and Sarkis (2020) and Pandya and Kumar (2023). A study by Bai et al. (2020) disclosed that mobile technology has the greatest impact on sustainability across all sectors. In contrast, research by Pandya and Kumar (2023) identified AI, BDA, and IoT as the primary drivers of sustainability.

The findings acknowledged AI (0.24577), MB (0.21517), and BC (0.19274) as the upper-level technological advancements accelerating the shift towards sustainability. These innovative technologies have been crucial for improving consumer interactions, modernizing banking operations, and safeguarding the transparency of financial transactions. For instance, the omnipresent impact of AI has transformed the banking industry by upending established norms and fusing conventional approaches with myriad innovations. Its incorporation can enhance lending, credit evaluation, cybersecurity, and regulatory compliance, and pave the way for sustainable development (Ozili, 2021). The predominant weightage of AI directly supports RQ1, which seeks to determine the technologies with the greatest potential to drive sustainability results. As we can see, AI has approximately five times the impact as RegTech, BDA, and IoT. On the other hand, MB has optimized every facet of financial transactions for corporate and individual customers. Due to the availability of traceability solutions, expanded financial inclusion, decreased carbon footprint, and cashless transactional procedures, mobile wallets have elevated the sustainability quotient of payment services. Thus, MB directly enhances both social and environmental dimensions, explaining around twice as much of the impact of CC in the hybrid MCDM study. Furthermore, BC has garnered attention for its third position in influencing sustainability mechanisms. The United Nations acknowledges this safe decentralized ledger technology as a revolutionary force for improving microfinance and remittance applications for rural smallholder farmers to promote sustainable development goals (Giungato et al., 2017). In response to RQ2 regarding the factors driving higher prioritization, the concentration of BC indicates a shift in banking towards authentication technologies that facilitate regulatory reporting, greenhouse gas tracking, and inexpensive cross-border transactions.

Expert opinions have positioned CC (0.09554), RPA (0.0948), and BDA (0.05979) as the mid-tier imperative technology among a compendium of advanced technologies. Despite being a fundamental enabler for digital transformation, CC amplifies cybersecurity in the banking industry by implementing a multitude of robust security measures, for example, complex threat detection systems, cryptographic encryption, and stringent access control procedures that protect confidential financial information from unauthorized breaches. Within this context, cc is perceived as a relatively mature, efficiency-focused technology with indirect sustainability advantages, with moderate weight (around half of BC). RPA, on the other hand, eradicates operational redundancies and paper-based administrative tasks, but it doesn't accelerate the sustainability transition as aggressively as AI, BC, and MB.

However, BDA (0.05979), IoT (0.05221), and RegTech (0.04396) propel banking

sustainability, yet these technologies have been assigned lower weights in the current study. The lower score of BDA advocates that though it is essential for insight development but data-driven sustainability projects demand substantial institutional maturity, which many banks may not have accomplished till now. Nevertheless, IoT holds enormous potential for increasing operational efficiency through improved connectivity and data interchange, but its impact on sustainability in the banking sector may not be as apparent as in manufacturing or logistics. Lastly, RegTech solutions aren't designed to operate with sustainability mechanisms or strategies. Despite being necessary for compliance, it is seen as a supportive technology rather than a transformational one, making a bigger difference to regulatory efficiency than to the direct creation of sustainability value. A sustainable business model cannot be exclusively achieved by investing in digitization or improving existing offerings. Banks should reinforce medium-to long-term aggressive digitalization strategies, focusing on enhancing market capabilities, establishing infrastructure, and product enhancements. The digital transformation should be judiciously orchestrated and envisioned to excel in all three dimensions of sustainability.

This study proposes multiple theoretical ramifications. The study is one of the few attempts to assess and hierarchize digital technologies within the sustainability paradigm in the banking industry; henceforth, its significance to the extant corpus of scholarly work cannot be understated. Additionally, the complementary nature of FDM and BWM reflects a clear picture of the rich tapestry of academic work and research contributions. Further, this study reviews diverse literature to give scholars and governors a broad understanding of digitalization and its potential to elevate social development, governance, and environmental quality. Additionally, the study showcases how these methodologies can be used across cross-domains in the technical sectors by connecting the dots between digitalization and sustainability literature. Further, the prioritization of sustainable digital technologies developed by the study provides a benchmark for future theoretical and empirical studies. The present technological landscape and evolving customer expectations for sustainable banking practices are both apparent in this prioritization. Lastly, this study can elucidate to scholars and practitioners how to conceptualize digitalization and sustainability as a single co-transformation rather than as two separate transformations running in parallel.

The Indian banking sector has addressed sustainability concerns with relative inertia (Kumar, 2019; Kumar & Prakash, 2020). Under such circumstances, banking institutions can use this framework version to prioritize investment in sustainable digital technologies identified in the present study. The findings can be used as a standard for other financial institutions to assess the sustainability and technology portfolios progress. The deployment of prioritized technologies can enhance banks' ESG performance indicators, supporting stakeholder reporting and regulatory compliance. Moreover, the outcomes can also be used by policymakers to develop specific incentives, rules, or support initiatives that promote the development of high-impact, sustainable innovations. Additionally, the study's findings hold substantial significance for managers who want to proactively leverage digital technologies to integrate sustainability into their operations, thereby meeting the 2030 Sustainable Development Agenda. Besides, top managers should encourage involvement in technology-enabled initiatives that promote sustainability and effective procedures by sharing knowledge and resources with staff members. These implications can guide researchers, practitioners, and decision-makers in their quest for advancement in the area of digital transformation in the banking industry.

The banking industry is currently encountering a profound transformation spurred by the accelerated progression of digital technologies. This digital transformation has emerged as the cornerstone for accomplishing sustainability targets in today's intricate landscape. Our paper has delved into a novel methodology for recognizing digital technologies through SLR, by which expert panels could identify, weigh, and prioritize sustainable digital technologies in the banking domain through FDM and BWM. For this purpose, authors have identified nine digital technologies through SLR, namely AI, BC, BDA, API, MB, RPA, IoT, CC, and RegTech. After that, nine experts assigned linguistic expressions and ratings to the technologies. The outcomes of FDM witnessed a defuzzified value less than the threshold for API. Consequently, eight technologies underwent BWM analysis. Finally, the technologies were ranked as follows by the BWM results: C1(0.24577) > C4(0.21517) > C2(0.19274) > C7(0.09554) > C5(0.0948) > C3(0.05979) > C6(0.05221) > C8(0.04396). The amalgamation of quantitative and qualitative perspectives has augmented the robustness of the findings, providing a comprehensive grasp of the continuous digital transformation in the banking industry.

Despite making a significant contribution, this study exhibited a bias towards the perspective of bank managers and academic experts, thereby potentially limiting the broader perspective on the findings. To overcome subjectivity in expert opinions, results could be compared and contrasted with secondary data on dominant players making massive investments in digital technologies. Further, a larger group of expert decision-makers with more domain expertise may be considered to gauge sustainable digital technologies. It will aid in obtaining holistic results. Future studies could determine whether the crucial technologies discovered are all-inclusive in other technical contexts. Additionally, the literature search for this study was restricted to works published up until April 2024. Consequently, it is possible that recent studies on digital technologies for sustainability in the banking industry were overlooked. To reinforce and broaden the conclusions, future evaluations should integrate updated evidence. Nevertheless, BWM establishes a priority ranking without explaining the relationship between the variables. Consequently, DEMATEL could be deployed to segment banking sustainable digital technologies into cause-and-effect groups in future studies. Other techniques like AHP and TOPSIS can also be used for prioritizing digital technologies, and the results obtained may be compared with the present study outcomes. Analogous studies might be carried out in fields such as manufacturing operations, the IT sector, etc.

“All author(s) read and approved the final manuscript.” –