Antecedents of behavioural intention to adopt central bank digital currency (e₹): examining the user acceptance of digital innovation in India

Central bank digital currencies (CBDCs) have become the sovereign digital equivalents to cash and privately issued cryptocurrencies, transforming the world economy, including in fast-digitising economies like India, with real-time settlement, efficiency and transparency (Alora et al., 2023). As the world moves in that direction, in December 2022, the Reserve Bank of India (RBI) started testing pilot projects of the retail digital rupee (e₹) in several large cities such as Mumbai, New Delhi, Bengaluru, Bhubaneswar and Chandigarh (Press Information Bureau, 2023). Central bank digital currency (CBDC) is a digital sovereign currency issued by central banks, which is technologically efficient and has the legal characteristics of fiat money (Auer et al., 2022), and is distinctly different in its fundamentals from decentralised cryptocurrencies, which depend on institutional trust and central authority (Bank for International Settlements, 2023).

CBDCs are reacting to the falling cash use, the diffusion of digital payments and the danger of substitute digital currencies, but they are also designed to improve the security of payments, financial accessibility and transmission of monetary policy (Kiff et al., 2020). International projects are characterised by diverse goals, such as the e-CNY of China as an example of payment sovereignty (Jiang and Lucero, 2022), the e-krona of Sweden as an example of maintaining access to central bank money (Sveriges Riksbank, 2023) and the Sand Dollar of the Bahamas as an example of financial inclusion in isolated locations (Wright et al., 2022). By July 2025, 137 countries with 98% of worldwide gross domestic product (GDP) consider CBDCs, and some are in their developmental formative stages, and three have fully operational retail CBDCs (Atlantic Council GeoEconomics Centre, 2025), which implies the growing necessity of empirical studies on the determinants of adoption.

The well-developed digital payment system, spearheaded by unified payments interface (UPI), in India registered a ₹24.03 lakh crore transaction value in 18.39 billion transactions in June 2025 (Press Information Bureau, 2025). Although there was an initial success, daily digital rupee transactions dropped to approximately 100,000 in the middle of 2024, down by one million from December 2023, suggesting that adoption was not successful with its first users. Although the use of CBDCs has advantages, including resilience to payments and the ability to encompass a wider sample of the population (Arner et al., 2020; Tan, 2023), evidence on CBDC adoption in India remains scarce and largely confined to the early pilot phase. To address this gap, the current study investigates intention to adopt the digital rupee among digitally active consumers in the cities where the retail pilot was first deployed, who represent its most probable early adopters, through the prism of an extended unified theory of acceptance and use of technology (UTAUT) framework, considering the constructs of perceived security, effort expectancy, perceived trust and performance expectancy and backed up by previous research on the familiarity with UPI (Gupta et al., 2023), the level of public awareness (Raman et al., 2025) and perceived trust and privacy issues (Tronnier et al., 2022).

The researchers conducted a detailed questionnaire-based study in the cities where the digital rupee pilot program was first implemented in 2022, to answer two key study questions:

The following are the objectives of the study:

There are three contributions of the study. First, this offers empirical evidence about the digitally active consumers from five early pilot cities of retail digital rupee – Mumbai, New Delhi, Bengaluru, Bhubaneswar and Chandigarh – as likely first movers. Second, it expands the theory of technology adoption by adding the TPB-based attitude to the extended UTAUT model, where the attitude acts as a mediator between the technology adoption determinants and behavioural intention. Third, it provides hands-on guidance for policymakers, regulators and technology builders to build positive sentiments around the CBDC and enable greater digital financial inclusion in India.

A central bank digital currency is a digital currency issued and backed by a central bank, which is not privately issued and carries the full faith of the state and is also under monetary authority (Allen et al., 2022). CBDCs are typically thought of as wholesale or retail, and can be account-based or token-based on centralised or distributed ledgers (Auer et al., 2023; Genc and Takagi, 2024). They have grown worldwide due to the reduction in cash transactions, surge in digital payment systems, worries about private cryptocurrencies and the need to maintain monetary sovereignty (Alfar et al., 2023; Kiff et al., 2020). CBDCs are considered to be financial inclusion drivers, payment-resistant assets, interoperable and transparent and disintermediate banks to a certain extent (Dionysopoulos et al., 2024; Prodan et al., 2024), but also to be exposed to cybersecurity risks, privacy concerns, design complexity and the possibility of disintermediation. The empirical evidence shows that trust in the issuing authority, perceived usefulness, usability and regulatory clarity are key factors that promote successful adoption (Ozili, 2023).

The UTAUT combines elements from eight previous behavioural acceptance models and identifies four key factors of technology acceptance: social influence, facilitating conditions, performance expectancy and effort expectancy (Venkatesh et al., 2003). The framework was later extended to the consumer context to include hedonic motivation, price value and habit (Venkatesh et al., 2012), and a comprehensive meta-analysis further validated the framework and found that its explanatory power enhances when including constructs specific to the context in which the model is applied (Blut et al., 2022). In financial services, model-based studies of mobile payments, internet banking and fintech in emerging economies consistently identify performance expectancy, ease of use, trust and security as significant antecedents of adoption intention (Acosta-Prado et al., 2024; Islam et al., 2024; Qasim and Abu-Shanab, 2016; Tariq et al., 2024; Wang, 2025).

The original UTAUT model, although powerful in explanation, fails to include attitude and treats intention as a direct result of the key elements. The TPB holds that intention is determined by attitude, which is the individual's evaluation of performing a behaviour in a favourable or unfavourable way (Ajzen, 1991). Dwivedi et al. (2019) also demonstrated the improvement in the predictive validity when attitude was reintroduced into the UTAUT as an attitude-based determinant that typically affects intention through users' overall assessment of technology. This is a path of attitude and is especially relevant in the case of the digital Rupee, which remains a voluntary adoption and requires a person's judgement. This research, therefore, combines the attitude component of the TPB into the UTAUT, adds performance expectancy and effort expectancy as technology-related beliefs and places attitude as the intervening variable between these beliefs and behavioural intention, and includes perceived trust and perceived security as finance-specific risk-related determinants.

Building on the combined structure above, the study defines each determinant, reviews the major works examining its relationship with attitude and adoption intention and develops the corresponding hypotheses.

It is the perception of users that the system adequately protects their personal and transactional information. Research on technology acceptance model (TAM) and UTAUT models reveals a consistent finding of positive impacts of high security perceptions on trust, reduced risk and intentions to adopt mobile banking and digital payment systems, although the effects might differ based on demographic groups (Almaiah et al., 2023; Jafri et al., 2023; Oliveira et al., 2016; Yao et al., 2023). This study predicted that perceived security would significantly influence attitudes towards CBDC. Based on the preceding theoretical and empirical evidence, the following hypothesis is proposed:

Effort Expectancy (EE) refers to the perceived ease of using a technology and the extent to which it demands minimal effort (Davis, 1989). Ease of use, in addition to performance advantages, is found to be a key factor influencing adoption intentions, as confirmed in studies performed in the education sector, fintech and digital payment systems (Marchewka and Kostiwa, 2007; Granić, 2023; Sultana et al., 2023; Kiran and Vedala, 2025). Hence, we hypothesised the following:

Perceived Trust (PT) implies the willingness of users to rely on a system under conditions of uncertainty and risk (Moorman et al., 1992). Trust is a key determinant of adoption intentions in financial technology settings. Recent findings show that higher levels of trust strongly strengthen users' intentions towards using fintech and robo-advisory services, especially in emerging markets where institutional trust is a prerequisite (Saeed et al., 2024; Bashir et al., 2025; Appiah and Agblewornu, 2025). So, the hypothesis is as follows:

Performance Expectancy (PE) is the assumption by an individual that using a technology would improve his or her ability to perform a task or to achieve a desired outcome (Venkatesh et al., 2003; Zhou et al., 2010). Previously, PE has been shown to be a strong predictor of adoption behaviour in the digital finance sector. The empirical study of the adoption of CBDC, UPI and fintech in India, Bangladesh and China (Sultana et al., 2023; Kurniasari et al., 2023; Jacob et al., 2024; Kaur et al., 2025; Wei et al., 2025) suggests that the perceived efficiency, convenience and performance benefits significantly affect behavioural intention. Therefore, we hypothesised that:

Attitude (AT) is defined as the positive or negative evaluation of using a technology (Ajzen, 1991; Lee, 2009). Attitude shows consistent predictive value for adoption in the context of financial innovation (FI). Empirical research in Nigeria, Vietnam, Ghana, China and CBDC contexts attests that those positive attitudes, conditioned by perceived usefulness, usability, performance value, convenience and trialability, exert a significant effect on fintech and CBDC adoption intentions (Ngo and Nguyen, 2022; Krah et al., 2024; Wei et al., 2025; Amin et al., 2025).

Behavioural Intention (BI) is an individual's readiness to use a technology and is a powerful antecedent of actual use (Ajzen, 1991). UTAUT- and TPB-based studies indicate that performance expectancy, effort expectancy, habit, perceived value, trust, compatibility, social influence and facilitating conditions significantly affect behavioural intention towards digital payments and mobile wallets across diverse settings, including the rural economy and community-based finance (Ong et al., 2023; Meiranto et al., 2024; Hasan et al., 2024; Khan and Abideen, 2023; Wauk et al., 2025; Chipeta and Lipunga, 2025). Hence, we hypothesised the following:

Effort expectancy, performance expectancy, perceived trust and perceived security are expected to have an indirect impact on behavioural intention through attitude. Based on TPB, the intention is dependent on overall behavioural evaluation of the individual, which makes attitude the mediating variable between these determinants and digital rupee adoption intention (Dwivedi et al., 2019).

Despite the increasing research interest in CBDC, there are three gaps. First, little empirical evidence exists on the behavioural intention towards the introduction of the retail digital rupee, especially from the cities where the pilot first began (Ozili, 2023; Kaur et al., 2025). Secondly, a majority of the adoption studies take intention as a direct predictor of adoption, whereas the mediation effect of attitudes is not widely explored in the context of CBDC adoption (Dwivedi et al., 2019). Thirdly, there is limited evidence on how trust and security influence perceptions of the state-issued digital currency. This study aims to fill these gaps by studying the antecedents of attitude and behavioural intention: perceived security, effort expectancy, perceived trust and performance expectancy.

Figure 1 shows the conceptual model. It positions the four determinants as exogenous constructs influencing attitude (H1 to H4), with attitude in turn influencing behavioural intention (H5) and H5a to H5d as the mediating effect.

This study used a descriptive cross-sectional research design to investigate the behavioural intention to adopt CBDC among digitally active consumers residing in five Indian cities, namely Mumbai, New Delhi, Bengaluru, Bhubaneswar and Chandigarh, where the RBI launched the retail CBDC pilot for the first time (Press Information Bureau, 2023). The choice of these cities was based on their strategic role in the initial retail digital rupee pilot and their established digital banking infrastructure. This study was conducted in the pilot phase of the retail digital rupee and investigated a pre-adoption behaviour construct, behaviour intention (Ajzen, 1991). Accordingly, the target population consisted of digitally active consumers in the pilot cities who knew about the concept of the digital rupee; these consumers are more likely to be those who will be the early adopters. This is in line with previous intention studies (Amin et al., 2025; Kaur et al., 2025).

In January and March 2025, the structured questionnaire was disseminated to the people through email and in person. The study adopted a purposive non-probability sampling method because it is a convenience and snowball sampling method, which is appropriate when the population to be studied is emerging and difficult to count (Kaur et al., 2024; Etikan et al., 2016; Sedgwick, 2013). The response was gathered from the theoretically relevant group of potential early adopters by restricting the sample to digitally active consumers who were aware of the digital rupee in their pilot cities. The study is designed to investigate theoretical relations, and therefore, generalisation of the study is possible as per variance-based partial least squares structural equation modelling (PLS-SEM) (Hair et al., 2017). The sample size exceeded the minimum requirement for adequate statistical power. Data were collected across five pilot cities through two channels (email and in-person), and the model demonstrated satisfactory predictive relevance. The questionnaire was administered to 450 individuals, and 341 completed responses were received, resulting in a response rate of 75.78%. The hypothesised relationships were subsequently assessed using a bootstrapping procedure with 5,000 subsamples, as recommended by Hair et al. (2020). The total number of relevant responses used in the Structural Equation Modelling (SEM) sample was 268 (Table 1) after removing 73 incomplete or incorrect responses. This questionnaire was completed with 21 items that surpassed the necessary subject-to-item ratio of 10:1 (Kline, 2023).

Data were gathered in English, which is the language they were most comfortable with when using digital technologies. The items were found to be clear and reliable in a pilot study of 50 subjects, with Cronbach's alpha values of more than 0.70 (Hair et al., 2017). Participation was voluntary and individuals were told about the study and its confidentiality.

The final number of participants was 268, with women and men evenly distributed to avoid gender bias (51.11% male and 48.89% female) (Windasari et al., 2022). Most respondents (63.8%) were postgraduates, while 30.6% were undergraduates. Of those, 46.64% were employed. The majority (83.56%) belonged to the middle-income group with less than ₹50,000 monthly income. This profile is a socioeconomically and digitally literate population segment which should be well suited for the study of early behavioural intention for CBDC.

To assess the ideas presented in the theoretical framework, a structured questionnaire was adopted. The instrument involved validated scales modified from previously published studies, with appropriate changes to reflect the context of CBDC deployment in India. The questionnaire has two primary elements. Section 1 comprised the demographic information of the respondents, while Section 2 included 21 items measuring significant aspects on a five-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree).

Perceived security was captured in a modified version of AlHassan et al. (2025) scale, while effort expectancy and performance expectancy were measured using UTAUT-based scales (Venkatesh et al., 2003; Dendrinos and Spais, 2023; Amnas et al., 2023). Measures of perceived trust, attitude and behavioural intention were based on previously validated scales in the field of FinTech adoption (Parajuli et al., 2024; Sharma et al., 2025). The results of a pilot test yielded satisfactory reliability (Hair et al., 2017). The WebPower was used for normality, and PLS-SEM was used to analyse data (Nitzl et al., 2016).

Consequently, the primary analysis of the structural model was done through SmartPLS 4 (v.4.1.0.2). Like any other tools based on variance as SEM, SmartPLS is preferred to be used with non-normally distributed data and exploratory model testing (Hair et al., 2017; Ringle et al., 2015). Also, statistical package for the social sciences (SPSS) v23 was used to generate the descriptive statistics of respondents' demographic characteristics.

To determine non-response bias, Armstrong and Overton (1977) used a paired-samples t-test of 50 early and late respondents. The absence of any significant differences indicated the absence of non-response bias (Lin and Schaeffer, 1995). Collinearity variance inflation factors (VIFs): Full collinearity VIFs were used in assessing common method bias in SmartPLS 4. The values of all VIFs were less than 3.3, which means that the model does not contain a significant common method bias (Kock, 2015).

Table 2 shows the assessment of the measurement model using SmartPLS 4. All loadings were above 0.60, which indicates the reliability of the indicators (Hair et al., 2017). The loadings for perceived security were between 0.660 and 0.800, performance expectancy between 0.768 and 0.872, attitude was 0.789–0.827 and behavioural intention was 0.862–0.932. The internal consistency was confirmed by applying Cronbach's alpha and composite reliability, which are all above 0.70, ranging from 0.709 to 0.888 (Hair et al., 2017). The composite reliability estimates ranged from 0.817 to 0.931, which was considered satisfactory. The values of average variance extracted (AVE) (0.530–0.817) were found to be in excess of the minimum value of 0.50 (Fornell and Larcker, 1981), which confirmed the convergent validity. Given that, the reliability and validity of the measurement model would be reliable.

The study's discriminant validity was analysed using the heterotrait-monotrait ratio (HTMT) ratio and the Fornell-Larcker criterion. Ringle et al. (2023) defined acceptable HTMT values as those less than 0.90, and all constructs in this investigation met this criterion, demonstrating varied uniqueness (Table 3). The Fornell-Larcker test confirmed this (Table 4), as the square root of AVE for each construct exceeded the inter-construct correlations (Fornell and Larcker, 1981). Together, these findings provide appropriate discriminant validity (Hair et al., 2020).

Table 5 shows how the structural model was evaluated using the bootstrapping approach with 5,000 subsamples, according to Hair et al.'s (2020) criteria. It shows the outcomes with the help of the path coefficients (β), t-values, permutation values (p-values) and VIFs. The outcomes of the study show that the constructs have a number of statistically noteworthy correlations.

Effort Expectancy (EE) has a substantial positive correlation with Attitude (AT) (β = 0.240, t = 3.302, p < 0.001), confirming H2. Perceived Trust (PT) correlated positively with Attitude (β = 0.190, t = 3.687, p < 0.001), supporting H3. Performance Expectancy (PE) was the strongest predictor of Attitude (β = 0.401, t = 5.682, p < 0.001), confirming H4. The pathway from Perceived Security (PS) to Attitude (AT) was not significant (β = 0.084, t = 1.516, p = 0.130); So, H1 was rejected.

The research model in Figure 2 explained about 62.2% of the variance in predicting attitude (R² = 0.622) and 47.7% in behavioural intention (R² = 0.477). Attitude and behavioural intention adjusted R2 were 0.615 and 0.475, respectively, which entails a high level of explanatory power (Hair et al., 2016). Moreover, both variables had predictive relevance (Q²) values that were above a value of 0, as indicated by Chin (1998), which were Q² = 0.604 for Attitude and Q² = 0.488 for Behavioural Intention, indicating the right predictive power of the model.

SmartPLS bootstrapping with 5,000 subsamples was used for mediation analysis (Nitzl et al., 2016). Table 6 shows that performance expectancy, perceived trust and effort expectancy significantly impacted attitude, which in turn significantly impacted behavioural intention, to support H5b, H5c and H5d. The indirect effect of perceived security was, however, negligible, and H5a is rejected.

The findings support the proposed integrated framework by showing that consumers' behavioural intention to adopt the digital rupee is shaped mainly through attitude (Ajzen, 1991; Dwivedi et al., 2019). This indicates that adoption of CBDC is not driven merely by awareness of the service, but by the overall evaluation users form on the basis of the usefulness, ease of use, trust and perceived security.

Performance expectancy also came out as a significant antecedent to attitude, revealing that consumers' attitude towards digital rupee is a positive one when they believe it to be useful, efficient and advantageous in the transaction process. The digital rupee, in India's established digital payment landscape which involves platforms such as UPI and mobile wallets (Gupta et al., 2023), should provide tangible operational benefits. This is in line with recent studies in CBDC (Venkatesh et al., 2003; Kaur et al., 2025; Sultana et al., 2023).

Effort expectancy was also found to influence attitude, and ease of use is still relevant during the initial adoption of CBDC (Granić, 2023; Kiran and Vedala, 2025). In addition, the digital rupee is not widely known, and a user-friendly and familiar experience can help lower consumer hesitation and increase acceptance. Likewise, perceived trust was a significant factor, as it underscores the importance of system confidence, issuer trust and the reliability of digital transactions in the adoption of FI from governments (Saeed et al., 2024; Appiah and Agblewornu, 2025).

Interestingly, the perceived security did not appear in the list of significant attitudinal drivers, unlike in the case of private digital-payment settings (Almaiah et al., 2023; Yao et al., 2023). This is one of the reasons why it can be assumed that users consider security to be a prerequisite standard of a CBDC.

Last, the attitude was found to have a strong influence on the behavioural intention, consistent with the TPB (Ajzen, 1991) and previous evidence of attitude being a key determinant of fintech adoption (Marakarkandy et al., 2017; Krah et al., 2024). The mediated effects are significant and demonstrate that attitude is a mediator between performance expectancy, effort expectancy, perceived trust and the behavioural intention to adopt the digital rupee. In summary, the study recommends that the penetration of the digital rupee can be enhanced by promoting its utility, ease of use and the confidence of institutions.

This study used the UTAUT and the TPB to explore the determinants of consumers' intention to use digital rupee (e₹) in India. The findings reveal that, based on the responses obtained from the digitally active consumers in the retail pilot locations, the variables “performance expectancy”, “effort expectancy” and “perceived trust” were found to have a significant impact on the attitude towards digital rupee. Attitude, in turn, serves as an important psychological mechanism linking these perceptions to behavioural intention.

Results show that digital rupee adoption is not just about being aware of it, but about the perceived utility, usability and trustworthiness of the digital currency by consumers. Interestingly, perceived security did not appear as an independent attitude determinant, meaning that security might be one of the factors that users may view as a property of a currency issued by the central bank and feel it is not a separate motivation.

The study adds to the existing literature on adoption of CBDC by enriching it with a theoretical understanding of the acceptance of digital rupee in an emerging-market setting and offering practical guidance to policymakers, regulators and financial institutions.

This study provides theoretical and practical implications to understand Indian consumers' acceptance of the digital rupee (e₹). It extends the UTAUT to add attitude from the TPB as a mediator between the adoption determinants and behavioural intention, theoretically. This allows for a more complete understanding of digital rupee adoption as it traces the consumers' attitude toward the adoption of the digital rupee into their behavioural intentions toward the digital rupee (Ajzen, 1991; Dwivedi et al., 2019).

The study also fits into the burgeoning body of literature on CBDCs by adding perceived trust and perceived security to the constructs of performance expectancy and effort expectancy. Unlike private digital-payment systems, the digital rupee is a sovereign digital currency issued by the central bank, which means users' trust and security judgements are shaped by confidence in a state institution rather than in a private technology provider. The results indicate that the factors of usefulness, ease of use and trust drive positive attitudes toward CBDC. The non-significant role of perceived security also suggests that security could be a characteristic of a central bank-issued currency that users may assume as part of the product, rather than something they evaluate independently.

In the practical aspect of things, the findings provide guidance to various digital rupee stakeholders, including policymakers and regulators, banks and technology companies that are part of the larger digital rupee rollout. Communication strategies should emphasise the ease of transactions, reliability, immediate settlement, low cost and potential offline usability. The digital rupee ecosystem must also make onboarding easy and wallets user-friendly with intuitive instructions for transactions. In general, to achieve greater adoption, the utility, usability and trust in institutions and effective merchant acquisition will be critical.

The research limitations are as follows. First, the data are collected in the five pilot cities, so the results may not apply to rural areas, Tier 2 or Tier 3 cities, and the sample is skewed towards younger, educated and digitally fluent individuals, who are early adopters, and not the general population. Secondly, the study used non-probabilistic convenience and snowball sampling techniques as there is no sampling frame available for the small population of digital rupee-aware consumers. The sample is not a statistical cross-section of all Indian consumers. However, the robust and positive predictive importance of the model and the consistency of the results with other scenarios suggest that the found relationships are generalisable to the realistic adopter base of digitally active consumers. Probability-based replication across rural, older and less digitally literate segments would further confirm the broader applicability of the results. Third, the cross-sectional methodology captures the user understanding at a single point in time and may not reflect shifts in attitudes as the digital rupee ecosystem evolves. Finally, the model restricted itself to a set of determinants and did not include other constructs that could be of interest, such as perceived risk, social influence, technology anxiety and government support.

Future studies could examine these limitations in a number of ways. This study's findings may be extended by using larger and more diverse sample populations (such as older adults, rural residents and less digitally literate users) and by using probability sampling if possible. Conducting a longitudinal study would be beneficial to assess the dynamics of intention to use and actual use of the digital currency over time, and adding other constructs and moderators like age, income and digital experience would strengthen the model. Last but not least, comparative cross-country analysis would offer relevant insights into CBDC adoption in varying regulatory and cultural environments in emerging economies.

This article does not contain any studies involving human participants or animals performed by the authors. Informed consent was not required.

I certify that this research work is original and has been prepared solely by the author. No artificial intelligence (AI) tools were used in the creation, writing, or structuring of the research content.