AI and Automation and Hassle-Free Bookkeeping: Testing Mediating Effect of User Acceptance and Moderating Impact of Digital Trust in the UK SME Retail and Service Sectors

1. INTRODUCTION

In every industry, Artificial Intelligence (AI) and automation are transforming the way business activities are performed, drastically changing conventional work processes. AI is a term used to describe a subfield of computer science that uses advanced analytical and logic-informed solutions, including machine learning, to process events, make automated decisions, and act in ways that historically necessitated an act of human intelligence (Stryker & Kavlakoglu, 2024). Automation is the application of technology to carry out pre-set, rule-based operations with minimal human interaction, thereby minimising repetitive human workload (Restrepo, 2024). These technologies have grown exponentially over the last ten years and have been applied in business functions such as accounting and finance.

For the AI market, it is predicted that it can grow exponentially as corporations persist with digital transformation plans and is projected to reach up to $4.8 trillion across the globe by 2033 (UN Trade and Development, 2025). Moreover, the accounting AI market is projected to be $50.29 billion by 2030, with automated bookkeeping software contributing to most of these increments (Prarthanna & Vadivel, 2025). AI and automation tools can be used to significantly streamline complicated business operations and increase operational efficiency in both the service and retail sectors by making it possible to handle real-time data processing, reduce errors, predictively insightful, and intelligent to perform various tasks (Dalsaniya & Patel, 2022; Kitsantas et al., 2024). Moreover, a synthesis of various industry reports revealed that AI-based automation can reduce 90% of manual data entry errors and speed up the transaction process by 50% in accounting (Eziefule, 2022). Taken together, these trends highlight the revolutionary role of AI and automation in contemporary business activities and the technological context of examining their effects on bookkeeping practices.

Theoretically, to comprehend the impact of AI and automation on bookkeeping results, it is necessary to base studies on technology acceptance and user behaviour models. The Technology Acceptance Model (TAM) assumes that perceptions of usefulness and perceived ease of use affect users’ attitudes toward technology which in turn impacts their intentions and actual adoption habits (Mogaji et al., 2024). Within TAM, trust in technology and acceptance of the person are organisational and individual factors that determine whether technological innovations are adopted and put into effective use. Furthermore, similar dynamics have been studied in other countries through empirical research. As an illustration, a study on the retail industry investigated the impact of AI technologies on the productivity of firms in online stores and found that the adoption of AI had a significant effect on sales conversion and operational efficiency due to the decrease in friction in online processes (Fang et al., 2025). Similarly, empirical research in service-based small enterprises has shown that the automation of routine administrative operations can enhance operational efficiency and decision-making capacity, although it is not based in the UK (Peláez & Álvarez, 2025). These external results suggest preliminary evidence that AI and automation can influence bookkeeping performance and show that studies need to be localised to consider specific socioeconomic and regulatory factors.

Small and medium-sized enterprises (SMEs) are the core of the UK economy, as they represent more than 99.8% of the total number of businesses in the country, which is imperative to the national economic performance and employment (GOV.UK, 2025). Although these companies are economically significant, they usually face manual and time-consuming challenges, especially in administrative areas such as bookkeeping and financial reporting (Brooks, 2023). Manual bookkeeping systems are labour-intensive, subject to human error, and distract management focus on the main entrepreneurial operations, which form operational bottlenecks and have the potential to restrict growth and competitiveness (Stevens, 2024). The adoption of technology in UK SMEs is not evenly distributed, with a significant number of SMEs still lacking the use of basic digital tools and well-organised plans to implement advanced technologies in their operations. The main obstacles are inaccessibility to finance to invest in technology, inadequate technical capabilities of small teams, and scepticism about the gains of digital innovation (GOV.UK, 2025). These restrictions do not just reduce the productivity but also increase the digital divide between companies who are technologically ready, and those who are trailing behind. Thus, the presence of large barriers to technological change and resource constraints characterizes the SME environment, given the potential of technological change is immense.

Therefore, the research aims to examine the role of AI and automation in supporting hassle-free bookkeeping in SMEs, in the UK retail and service industry, the mediating role of user acceptance, and the moderating role of digital trust. The current research adds to the theory by developing a TAM-inspired model within the SME environment, which is characterized by scarce resources and the variability of technological readiness. Methodologically, it pays attention to research gaps existing in the existing literature through introducing some insidious constructs, including digital trust, that demonstrate how much users are confident in the accuracy of technology, data safety, and ethical processing of financial data as a component of the acceptance and use model. In practice, the study is useful to SME owners, technology providers, and policymakers who would wish to make businesses more resilient, efficient in managing their bookkeeping operations, and use more digital in the future.

This paper gives a thorough description of the theoretical framework underpinning this study in the latter sections, followed by the generation of hypotheses based on the existing literature. The methodology section provides every step that has been taken to complete this study. The results and discussion are presented in the designated sections, followed by the conclusion, addressing the key findings, future directions, and limitations of this study.

2. LITERATURE REVIEW

2.1. Theoretical Framework

In this study, the theoretical approach used was TAM. TAM was initially theorised by Davis, and it describes the ways users become acquainted with and adopt new technologies, where the behavioural intention toward technology adoption is determined by perceived usefulness (PU) and perceived ease of use (PEOU) (Mogaji et al., 2024). Here, PU is having belief in a system that it improves job performance and makes a person or work better or more efficient, whereas PEOU is the belief that the system is effortless such as it is easy to learn and understand. User acceptance is dependent on these constructs, hence affecting actual use.

Empirical research has recently been applied to assess elements such as technology adoption in the SME context across industries. To illustrate this, a survey on sustainable technology adoption of SMEs in service (beauty, health, and wellness) industries found PEOU and social influence to be significant predictors of technology adoption, which endorses the relevance of TAM in service-based small firms (Qing et al., 2025). Likewise, a Lebanese study examined the long-term TAM in the adoption of accounting information systems (AIS) by SMEs (Dallal & Sankari, 2025). PEU, trust, and security were found to positively influence the acceptance and use of AIS, indicating that TAM can be used in financial technology applications in bookkeeping activities. Furthermore, the TAMs main constructs, PU, and PEOU, have been found to spur adoption intentions and performance outcomes in retail and business settings (Ojeka-John et al., 2025). Taken together such studies show the applicability of the TAM in the study of the role of individual and organisational perceptions in shaping technology adoption behaviours within retail and service SME settings, which provides a solid theoretical basis to the current research.

2.2. Key Constructs

Bookkeeping is the process by which financial transactions of a business are systematically captured, organised and tracked. Adequate bookkeeping can ensure effective reporting, minimize errors and enable companies to observe the flow of money and their profitability. This is especially necessary for SMEs, which usually have narrow margins and limited resources (Adela et al., 2024). In the context of this study, hassle-free bookkeeping refers to AI- and automation-enabled streamlined financial recording processes that improve accuracy, save time, and reduce administrative burdens.

In addition, user acceptance is the level at which individuals are ready to use a new technology to perform tasks (Wisdom Library, 2024). When it comes to AI and automation, user acceptance decides whether employees or managers would use these tools to bookkeep. This is significant because successful utilisation, productivity increase, and return on investment in technologies are achieved with high acceptance (Saira et al., 2025). However, when user acceptance is weak, even the most sophisticated tools are not used properly, which leads to the continued presence of manual work, low efficiency, and organisational resistance to change. Digital trust refers to the trust that users have in the reliable functioning, security, and ethical operations of digital systems (World Economic Forum, 2024). It affects how users are willing to share their data, assign some tasks to automated systems, and use technology to perform important business activities (Pietrzak & Takala, 2021). Digital trust is essential because it minimises risk perception and promotes the wider use of technologies. A lack of digital trust can make users fear data loss, information misuse, or algorithm failures, which can make them hesitate to use digital tools or even reject them altogether (Hasija & Esper, 2022). These constructs, combined, influence the successful adoption of AI and automation within the booking systems of SMEs, as well as their final performance. Having a clear picture of every construct assists in explaining the nature of technology behaviour and informs how to advance the digital transformation of business settings.

2.3. Hypothesis Development

(Abidemi, 2024) conducted a study to examine the impact of automation technologies on streamlining business processes and improving business performance among SMEs across different industries, such as retail and services. It was found that those SMEs that adopted automation tools reported a major shift of up to 30% in operational efficiency, a significant 25% reduction in operational errors, and increased productivity, specifically in automated operations like data entry, inventory management, and bookkeeping (Abidemi, 2024). The strength of this study is that it uses sector-specific data, which proves the possibility to document the improvement in performance with the implementation of automation in the real SMEs. The findings offer practical evidence that automation reduces the occurrence of repetitive bookkeeping processes and enhances overall productivity of the businesses. However, a weakness is that it only demonstrates efficiency improvements, but does not directly indicate a bookkeeping-specific effect, and does not directly study retail or service SMEs or account for variation in firm size and technology sophistication, which may limit external validity.

Expanding on these results, the effect of AI adoption in accounting has been also researched to present a complement of the automation-oriented studies. To illustrate, an empirical study on the impact of the adoption of AI on accounting efficiency, data quality, and fraud detection in Lebanese firms showed that AI tools have a significant impact on financial data effectiveness and quality (Bou & Jabbour, 2024). This technology improves fraud detection in financial records and has a positive impact on work activities and skill requirements among accounting teams. The theoretical merit of this study is that AI can be quantitatively supported to improve efficiency and data reliability, which are essential elements of hassle-free bookkeeping, based on a sufficiently large sample of accounting practitioners. However, its major limitation is that it addressed a specific country setting (Lebanon) and not the UK or other industry settings beyond retail and services, which may restrict the external validity of the findings to UK SMEs.

Extending the geographic and contextual scope, (Elom & Atah .2025) conducted a study of SMEs in Cross River State (Nigeria). They studied the effect of computerised accounting systems, such as automation in recording transactions, generating financial reports, and tracking income and expenses on financial performance. It was discovered that automatic recording of transactions and income/expense tracking had a significant and positive influence on financial performance, that is, it enhanced the accuracy and speed of financial data processing. The study is robust as it concentrates on the features of an automated financial system directly correlated with the results of bookkeeping and uses statistical analysis to prove any impact. However, its weakness is that it is geographically located outside the UK and uses cross-sectional survey data, which restricts the ability to cause and generalise. Together, these studies give convergent information that AI and automation technologies can enhance performance of operations, efficiency in accounting and financial data quality processing in small and medium enterprises. Thus, this study proposes Hypothesis 1:

H1: AI and automation have a significantly positive impact on hassle-free bookkeeping among small and medium-sized enterprises in the UK retail and service sectors.

(Relifra et al., 2025) applied an integrated TAM-UTAUT model to conduct an empirical study examining the impact of AI adoption determinants on functional business performance among micro-, small-, and medium-sized enterprises (MSMEs). In particular, it was discovered that behavioural intention (a key user acceptance measure) was a significant mediator between the determinants of AI adoption (e.g., performance expectancy and ease of use) and business performance outcomes (including in areas such as financial operations and operational processes). Behavioural intention reported a substantial variance in functional performance, and user acceptance was observed to be the factor that facilitates adoption drivers in affecting business outcomes in SMEs. This experiment provides the first direct correlation between user acceptance and performance results, which uniquely empowers mediation pathways. However, it is placed in a developing economy and does not specifically look at bookkeeping but at general performance.

Based on this understanding of user acceptance as a mediator, further studies have examined the operation of these constructs in specific accounting technologies. A study centred on SMEs in Denpasar examined the implementation of cloud accounting systems (automation tools), which are closely linked to AI bookkeeping (Putri et al., 2025). The researchers concluded that perceived ease of use and perceived usefulness were important in increasing the system’s effectiveness, and the system’s performance mediated the correlation between user perceptions and the firm’s performance. This is congruent with the fact that user acceptance (measured by perceived usefulness and ease of use) determines the translation of technology into results. Its strength lies in providing first-hand evidence in the SME accounting setting, explaining how the factors of acceptance can be used to translate technology features into performance. However, it is limited, as although relatable, it concentrates on cloud accounting as opposed to specific AI/automation tools and is located geographically outside of the UK.

In addition to the evidence from cloud accounting systems, the literature has further studied how SME technology adoption (AI) unfolds over time and provides insights into broader patterns of system use. A long-term study of the TAM regarding the adoption of Accounting Information Systems (AIS) in Lebanon concluded that perceived usefulness and trust positively impact AIS adoption, and acceptance constructs are also related to actual usage (Dallal & Sankari, 2025). Although this study did not test a complete mediation model involving performance outcomes, it affirmed that acceptance factors are mediators in the relationship between system characteristics and technology adoption behaviour. Strength: It expands TAM using variables applicable to accounting systems and is conceptually consistent with user acceptance in bookkeeping situations. Thus, this study proposes Hypothesis 2:

H2: User acceptance mediates the relationship between AI and automation and hassle-free bookkeeping among SMEs in the UK retail and service sectors.

One study that investigated the acceptance of AI service robots among customers in hospitality settings considered the interaction between trust and technology acceptance constructs in influencing the intention to adopt AI-enabled service automation (Chi et al., 2023). The results revealed that trust in AI is a strong predictor of the intention to use AI service robots, and higher trust levels are correlated with the acceptance of AI tools in service delivery. These findings support the idea that trust moderates the influence of perceptions of AI performance on user acceptance. The strength of this study is based on its service industry background and empirical consideration of trust as a structural influence that correlates with technology acceptance behavioural intentions in hospitality automation. Nevertheless, its greatest weakness is the lack of statistical testing of trust as a mediator between automation features and acceptance, and its sample was also limited to the consumer population and not to SME employees, which limits its generalisability to the organisational setting.

Extending the conceptual role of trust in user acceptance, research has also explored its influence specifically within SMEs, providing a more organizationally relevant perspective. For example, (Subhani et al., 2023) researched the impact of trust in technology on the adoption behaviour of SMEs in the ERP adoption setting. The researchers discovered that trust played a bigger role in enhancing the connection between the effort-related facets of technology (such as effort expectancy and task technology fit) and actual adoption behaviours. This is indicative of the fact that the higher the level of trust in systems, the greater the positive influence of technology’s characteristics on adoption. This provides conceptual evidence that trust is a moderator in technology acceptance models. The merit of this study is its SME focus and moderation analysis, which is consistent with the concept that trust improves the technology acceptance relationship. Nonetheless, its shortcoming is its particular concentration on ERP systems and not on AI/automation bookkeeping aids and geographic locations beyond the UK, which makes it less directly applicable.

Complementing this study, recent research has integrated trust with other key TAM constructs in SMEs to capture its role in AI adoption more comprehensively. (Popa et al., 2025) conducted a PLS-SEM study on the drivers of AI adoption among SMEs and included factors such as AI knowledge, perceived usefulness, ease of use, and trust in a comprehensive TAM. The results indicate that AI system trust has a strong positive impact on the behavioural intention to use AI tools, and that transparency and personalised experiences are significant determinants of trust that play a central role in the intention to use AI by users. Although the moderating effect was not directly tested in the study, it proved that trust positively influences users’ perceptions of AI benefits and usability, which means that the relationship between technology attributes (e.g., perceived usefulness) and acceptance results can be intensified by increasing trust. Thus, this study proposes Hypothesis 3:

H3: Digital trust positively moderates the relationship between AI and automation and User Acceptance, such that the relationship is stronger when digital trust is high.

2.4. Research Gap

Although recent literature has implemented TAM to explore SME technology adoption, some critical gaps remain. The literature supports the applicability of perceived usefulness and perceived ease of use to influence adoption in service, retail, and accounting cases (Dallal & Sankari, 2025; Ojeka-John et al., 2025; Qing et al., 2025). However, a large part of this research looks at general performance outcomes or larger accounting systems as opposed to bookkeeping-specific outcomes, such as hassle-free bookkeeping. In addition, existing research on AI and automation in the accounting field mostly focuses on developing economies (for example, Lebanon and Nigeria), which does not allow contextual generalisation to UK SMEs (Bou & Jabbour, 2024; Elom & Atah, 2025). Although user acceptance has been identified as an intervening variable in technology-performance correlations, very few studies have directly tested this variable in the context of AI-based bookkeeping. Despite trust being proven to affect technology acceptance, there is a lack of research on its moderating effect on TAM-based AI bookkeeping adoption in SMEs. This study fills these gaps by discussing AI-based bookkeeping in UK retail and service SMEs using a mediated and moderated TAM model.

2.5. Conceptual Framework

Fig. (1) presents a pictorial and concise view of the generated hypotheses, showing the conceptual framework for this study, which has not been studied simultaneously in a single study. Fig. (1) conceptualises that AI and automation are directly associated with hassle-free. Moreover, user acceptance mediates the relationship between AI and automation and hassle-free bookkeeping, where digital trust moderates this relationship.