1 Introduction

Economists consider artificial intelligence (AI) as a “general-purpose technology” [1] that, will significantly increase economic production and change the workforce [1, 2]. The impact of this technology is expected to grow rapidly because it plays a vital role in business and the global economy; hence, it is forecasted that future economies will rely heavily on it. Different industries have been transformed by technological advancements and accounting is not left behind. With complicated accounting procedures, AI must be applied for simplification and streamlining. This study attempts to understand the probable effects of AI use in accounting procedures and its implications in Saudi Arabia among those working in the accounting profession. In response to the study question, ‘How do AI implementations affect the performance of accounting procedures?’ This study aimed to delineate the purpose, goals, budget, target audience, and anticipated outcomes of examining the potential impact of AI on accounting procedures. Through an examination of the present state of AI and a critical assessment of its merits and drawbacks in accounting procedures, our objective is to provide valuable perspectives on its implications. We are confident that this analysis will enable the accounting industry to proactively prepare for the potential impacts of AI adoption.

AI is expected to significantly impact future­ economies. In 2018, McKinsey estimated that AI could raise global GDP by 1.2% per year. By 2030, it could add approximately $13 trillion to the world economy [3]. AI has a huge potential to boost production and earnings. Advances in AI may result in rapid innovation and major technological changes. This could drastically alter existing industries and the overall economy. Although AI has the potential to improve the world, not all of its advantages may be equitable. The accounting sector has been significantly affected by AI, which has increased its efficiency and automated operations. AI increases production quality and productivity [4, 5].

Accounting used to be­ about being correct. Each pe­nny must be­ counted. Currently, AI can assist with accounting operations. AI is used for taxes, book che­cking, and money reporting. AI in accounting has many benefits that can be emphasized; this approach not only automates repetitive processes, but also increases precision and effectiveness. Advanced fraud-detection methods have also been developed to lower the possibility of financial loss. As expected, the integration of AI is changing the role of accounting professionals [6].

According to recent estimates, AI might quadruple the annual rate of global economic development by 2035 through “intelligent automation, labor and capital augmentation, and innovation diffusion.” As AI technologies and systems have the potential to significantly increase production and earnings, their influence is enormous [7]. AI developments are anticipated to usher in a previously unheard-of-era of rapid innovation and technical transformation, drastically altering the existing sectors and economies. Although AI has the potential to improve the world, not all of its advantages may be equitable. Any new technology has a number of hazards, but one of the most significant is that it may increase economic inequality. The influence of AI on the global economy increases significantly as technology develops and strengthens. This impacts practically every area of the global economy, including income inequality, economic growth, and unemployment rates [8].

The advent of intelligent automation has ushered in significant transformations in the domains of numerous conventional professions, including accounting [9]. This paradigm shift is a product of societal order’s initiative to introduce and implement innovative technologies. The accounting profession is considered to be one of the professions affected by society. In this context, social considerations do have a significant impact on the accounting profession[10]. Furthermore, the ACCA Global article highlights the importance of societal concerns in determining the direction of the accounting profession [11]. Organizations need quick and accurate financial information for their operations, which prompts them to use advanced technologies. The perspective commonly known as technological determinism posits that technology serves as the primary catalyst for societal transformation. This concept is frequently associated with Thorstein Veblen, an American sociologist who recognizes the causal relationship between technology and society [12]. Advocates of this theory contend that technological advancements, including communication technology and media, exert substantial influence on social change. They also argue that society is molded and influenced by these advancements. To remain pertinent, individuals must adapt to and keep pace with the latest technologies and innovations [12]. Thus, incorporating technological advancements, including AI, into accounting procedures is not merely an option but rather an unavoidable result of technological development and social organization. This development will contribute to increasing the quality and credibility of financial information by reducing human error, and this quality will also play a role in increasing the quality of decisions. [13] expected future of accounting is positively influenced by AI, as it has the potential to automate numerous tasks, provide reliable financial data, simplify complex accounting and auditing situations, and deliver precise and timely information to make informed decisions [14]. also suggests that a major issue in generating subpar accounting information is the inadequacy of internal control systems. Contemporary technology has aided businesses in addressing this weakness by enhancing their internal control systems.

Over the past few years, technological advancements have significantly changed how businesses operate and provide accounting information [5]. AI is one of the most significant developments in human thought and work. Undeniably, technological upgrades have drastically changed the course of business in modern times. These changes have influenced accounting, a fundamental business activity that involves systematic and comprehensive recording of financial transactions. Businesses are now using more effective methods than AI to deliver accounting services [4, 15]. This study explores the effects of AI technology on accounting procedures in Saudi Arabia. Specifically, the main goal of this study is to determine whether the employment of AI technologies has impacted accounting procedures in Saudi Arabia. Since the release of Saudi Vision 2030, Saudi Arabia has made bold moves to promote the future of its economy and innovation [16].

This study contributes to the wide exploration of the possible influence of AI on accounting procedures in Saudi Arabia. In particular, the research assesses the knowledge, perceptions, and practices of accountants towards the adoption of AI in accounting procedures, correlating its results with industry trends indicated in the literature review. The study used a stringent survey methodology that involved a structured questionnaire and partial least squares (PLS) technique for analysis to ensure holistic data gathering while placing more emphasis on convergent and discriminant validity. Significantly, this study effectively validated metrics, proving their distinctiveness and dependability in evaluating AI knowledge, attitudes, and practices. The transformative nature of AI is evident from strong support for hypotheses concerning different dimensions of accounting being affected by artificial intelligence supported with training, among other factors. Through a detailed mediation analysis, this study shows how exposure to or experience with AI interact with each other and jointly affect accounting outcomes, thus contributing to a wider comprehension of the various impacts of AI. The findings imply that AI engagement and impact serve as important intermediaries in these associations, implying that accounting results can be improved by increasing awareness and usage of AI alone. Furthermore, this study provides insights on AI implementations into accounting procedures for current or future researchers interested in the subject matter, as well as providing useful hints for businesses and policymakers in Saudi Arabia where they may need such information while planning for investment in technology within changing trends.

The following sections cover the literature that has already been written about the impact of AI adoption in accounting; our research methodology; the findings of the direct relationships between AI awareness, usage, and engagement with accounting procedures; the implications for practice, education, and policy; and a summary of the main conclusions and recommendations for further research.

2 Literature review and hypothesis development

AI is a branch of computer science in which machines accomplish functions, such as learning, reasoning, and decision-making, that normally require human intelligence. Various accounting fields such as auditing, financial accounting, management accounting, taxation, and governments have adopted AI. Apart from offering decision-support insights and continuous auditing capabilities, this technology can help accountants improve their operational efficiency levels while enhancing accuracy rates and fraud-detection mechanisms. On the contrary, it creates opportunities for job displacement, raises ethical considerations, and demands new knowledge and skills from professionals [15]. Furthermore, through real-time monitoring, AI enables assets and stocks to remain under constant supervision. This enhances the generation of real-time reports and significantly boosts financial forecast precision. AI can also assist in reducing fraud by implementing ongoing financial auditing procedures that verify organizations' adherence to both local and international regulations [15, 17].

The existing body of literature on AI and accounting is still growing, and includes both theoretical and practical aspects. For example, some studies have suggested methods for integrating AI into the accounting curriculum, such as incorporating supplementary readings and mandating that students employ or develop straightforward expert systems[18]. Other studies have explored the influence of AI on the accounting industry and the role of accountants in the digital era [see 19–24]. [25] found that there are mechanisms for replacing accounting tasks with AI technology in operation, but accountants still show some reluctance to adopt this technology. AI, a formidable and disruptive technology, may drastically change the accounting and financial industries. To completely comprehend and utilize the advantages and difficulties of this technique, additional studies and explorations are necessary. Accounting professionals must adjust to the shifting environment and seize opportunities presented by AI.

The elimination of repetitive tasks is one benefit of accounting. AI integration can automate data entry, invoicing, and financial reporting, saving accountants’ valuable time for complex and strategic work. This automation also significantly reduces the risk of errors and ensures the accuracy of financial information [26, 27]. Additionally, AI-powered accounting solutions can detect and prevent fraud using advanced techniques such as anomaly detection and machine learning-powered pattern recognition. As these fraud-detection tools become more sophisticated, the chance of detecting financial malpractice increases [28]. Although AI has significant advantages in the accounting industry, its implementation can lead to undeniable changes in job roles. Robots will not replace accountants; however, the required skills will change and must be addressed by accountants. New jobs, such as data analysts and IT professionals, will proliferate, and a combination of accounting knowledge and technical skills will be needed to adapt to the emerging paradigm. The effect of AI on the accounting industry is vast. While there are potential challenges in integrating AI technology into accounting processes, including data security and implementation costs, AI power will significantly shape the industry's future [6, 21].

Moreover, AI has the potential to impact accounting practices in several ways. AI can analyze extensive quantities of nonfinancial and financial data to extract insights and predict financial outcomes more accurately than humans can [29]. It can also detect patterns of anomalies that may indicate potential fraud, with AI models being able to identify up to 92 percent of financial statement fraud cases. Moreover, AI can increase the productivity of audit processes by up to 25%, and enhance the accuracy of accounting estimates and accruals. AI can automate data gathering, consolidation, and analysis to generate financial reports faster and more accurately, potentially reducing time and cost by 50%. However, challenges continue to exist regarding the need for high-quality data, transparency, security, and ethical considerations in employing AI for accounting.

The integration of AI within the accounting field has increased over the past few decades, particularly in relation to auditing functions [30]. This has led to a transition in research approaches from traditional methods to experimental and archival studies [31]. Moreover, the emphasis on integrating AI into managerial accounting systems is driven by the need for additional empirical research in this domain [32]. Despite concerns about job displacement, AI has the potential to enhance accountants’ work, especially when they acquire essential digital skills [33].

The application of AI in accounting has become a topic of extensive research. AI was found to have a broad impact on efficiency, productivity, and customer service in the finance sector [34]. However, the ethical issues concerning implementing AI need more study [30, 32]. Research is ongoing to understand the difficulties facing the adoption of AI and how it can improve current accounting practices [35, 36]. Despite the potential of AI, it has been reported that resistance in AI technology adoption by some accountants persists[25]. The intersection of AI technologies with accounting procedures is complicated and changes at a rapid pace. Accounting programs are of prime importance in terms of integrating relevant IT knowledge and skills [37]. There has been inadequate focus on a clear definition of Accounting Information Systems (AISs), which has limited research in this area [38, 39]. AI is an important emerging trend in auditing and accounting applications, with leading consultants actively participating in the development of AI tools for risk evaluation, among others [40]. However, the use of AIS is not prevalent in business schools, however [41]. There have been continuing calls for the expansion of AIS research in relation to management accounting and control [42]. Further clarification of the definition of AIS could also better advance this field [39].

The future of the accounting industry should embrace AI with open arms. One of the key advantages of AI is the increased mechanization it brings to the table. Repetitive, timeـــconsuming tasks for example data entry and statement of financial position reconciliation can be completed in minutes rather than hours. This means that accountants have more time to gather insightful data for their clients [43, 44]. Nevertheless, AI improves decision-making abilities in addition to automating procedures. This method can help clients achieve their business purpose and increase their financial reporting accuracy. Furthermore, to improve accounting procedures, AI can be combined with other technologies such as blockchain and cloud computing. For example, blockchain enables transactions to be recorded and validated without the need for a centralized authority, thereby improving efficiency and accuracy [45,46,47,48].

AI has significantly enhanced accounting procedures by providing exceptional capabilities that are not found in other technologies such as blockchain, big data, enterprise resource planning (ERP) systems, and blockchain. AI has raised the bar for automation, intelligence, and productivity in industry. AI-powered optical character recognition (OCR) and natural language processing (NLP) can automate data entry from invoices and receipts with high accuracy, thereby decreasing manual effort and errors, according to a case study by [49]. [50] showed that AI is capable of processing large amounts of data and surpassing conventional statistical models in predicting financial parameters such as revenue, cash flow, and distress risk. The AI frameworks [51] and [52] for identifying false financial statements and the AI-powered risk-management solution [53] for ongoing transaction monitoring show that AI is very good at detecting fraud. To save time and effort while maintaining thoroughness, a study by [54] suggests AI-based frameworks for automating risk assessment, data extraction, auditing processes, and compliance review. According to research by [55], cognitive process automation (CPA) has been successfully applied to complicated accounting operations like financial reporting, journal entry management, and account reconciliation. CPA automates decision-making processes with little to no human involvement. Even though large amounts of data are essential for data-driven decision-making, AI’s intelligence and flexibility surpass that of big data. Similarly, while ERP systems simplify processes, they are not AI flexible. Blockchain enhances security and transparency, but falls short of AI’s analytical and decision-making abilities. AI’s remarkable capabilities of AI have the potential to revolutionize accounting procedures and increase efficiency and data-driven decision-making. While other technologies complement accounting processes, AI is a transformative force that automates, analyzes, and optimizes workflows in unparalleled ways, as demonstrated by numerous research studies and case experiments.

The accounting industry is changing with the emergence of artificial intelligence, which can automate basic accounting tasks and enable accountants to assume managerial accounting roles [56]. By using artificial intelligence, accountants can save time and energy on data entry and analysis and focus more on explanations and recommendations for business decisions [57]. However, artificial intelligence also brings some challenges to the accounting profession, such as ethical, legal, and social issues, and the need for human supervision and judgment [58]. A case study from the global accounting firm KPMG demonstrates how AI can improve the productivity and quality of accounting services, such as audits, taxes, and consulting [59]. Artificial intelligence can also improve business and investment outcomes for accounting clients by providing faster and more accurate insights and recommendations [59].

To answer the question, ‘How does the use of AI implementations impact the performance of accounting processes?’ We do so by focusing on specific AI applications and placing this question in the context of Saudi Arabia itself, completing repetitive tasks and analyzing large amounts of data. ask this question. Massive amounts of data have revolutionized accounting [60]. The integration of AI into accounting has the potential to enable accountants to concentrate on more advanced tasks such as financial analysis and business advisory services, while ensuring precision and adherence to regulations. Nevertheless, its incorporation raises ethical concerns including job displacement, data privacy, algorithmic bias, and fraud detection. AI can automate routine tasks, potentially reducing the need for traditional accounting roles and necessitating workforce management and reskilling initiatives. It is crucial to implement stringent security measures and adhere to data-protection regulations when handling sensitive financial information. AI algorithms may perpetuate biases present in the training data, leading to unfair or discriminatory outcomes that necessitate continuous monitoring and adjustment. On the other hand, AI can bolster fraud detection mechanisms by identifying unusual patterns and anomalies in financial transactions, thus enhancing financial integrity and corporate responsibility [60].

The above-mentioned research and technological determinism theory show that the combination of artificial intelligence and accounting has brought fundamental changes to accounting practices, and opportunities and challenges coexist. This study hypothesizes that there is a relationship between the use of artificial intelligence and accounting practices in Saudi Arabia. This hypothesis suggests several relationships between AI adoption and accounting. Figure 1 shows the path analysis model created to examine the hypothesized relationships. The first hypothesis, H1, states that awareness and use of AI have a significant impact on financial transaction processes. H2 advocates a link between AI awareness and utilization and accountants' participation in AI. H3 focuses on the necessity of preparing accountants for alterations caused by artificial intelligence. The fourth and fifth Hypotheses H4 and H5 explore the impact of AI on financial processes, accounting, accuracy, and costs. The sixth to eighth hypotheses H6-H8 examine the impact of demographic factors on AI adoption of artificial intelligence. The ninth and tenth hypotheses, H9 and H10, address the relationship between AI adoption, readiness for change, and accounting efficiency. The thirteenth to fifteenth hypotheses H13-H15 introduce the concepts of indirect relationships and intermediaries to AI adoption. This study aims to improve the understanding of artificial intelligence integration in accounting.

Fig. 1
figure 1

Hypothesises and model measurement

Full size image

3 Research method

Saudi Arabia was chosen as the research setting has been made based on several noteworthy considerations that highlight its suitability for investigating the effects of AI on accounting practices. Saudi Arabia boasts a national strategy for the use of AI and digital transformation. The Saudi Authority for Data and AI’s National Data and AI Strategy aims to establish the Kingdom as a global leader among the few data-driven economies. This initiative is an important part of Saudi Arabia’s Vision 2030, a long-term strategy to reduce the country’s dependence on oil and diversify its economy [16]. With AI technology being used to automate tedious accounting activities, detect abnormalities, and improve financial analysis, the financial services sector in Saudi Arabia is undergoing substantial transition [61]. The requirement for accuracy and efficiency in financial reporting and management drives this increase [62]. Saudi Arabia’s unique cultural and regulatory landscape presents distinct challenges and considerations for AI adoption in accounting. For example, incorporating AI into Saudi enterprises’ accounting systems may influence workforce dynamics, skill requirements, and organizational structure as a whole [see 63–66]. Although AI cannot completely replace human knowledge, it is revolutionizing the accounting business by enabling experts to work more quickly and efficiently [5]. As a result, using Saudi Arabia as the empirical environment for this study not only makes it more relevant within the Saudi context but also establishes the foundation for further investigation and cross-cultural comparisons in the fields of accounting and AI.

This study explores how AI, a key driver in future economies, affects accounting procedures. This study focuses on the effects of AI implementation on accounting procedures in Saudi Arabia. We use a questionnaire to collect data from accounting practitioners and professionals in Saudi Arabia. This study utilized a quantitative methodology to address the research question, “ How does the utilization of AI affect the performance of accounting procedures?” This study’s approach was based on a survey. It combines, changes, and updates the theoretical framework from previous empirical investigations [see 67–79]. The questionnaire consisted of three parts. The first portion of the survey requested information on respondents' demographics. The following portion of the survey asked six questions that measured respondents' degree of knowledge, attitudes, and practices regarding AI in accounting. The third part contained ten questions to measure the impact of AI on accounting processes.

3.1 Instrument design and measurement

The survey is organized into six sections, each centered on specific facets pertaining to the influence of AI on accounting practices. In Sect. 1, participants were asked about their perceptions of AI's impact of AI on accounting for financial transaction processes. We presented four questions on the use of AI to enhance the accuracy and efficacy of recording transactions, classifying transactions, summarizing financial statements, and producing financial information for stakeholders. Responses were collected on a 5ـــpointـــLikert scale. In Sect. 2, we discuss three questions that focus on the larger impact of AI on accounting efficiency, accuracy, and costs. Participants were permitted to discuss their thoughts on how AI affects the effectiveness and efficiency of accounting procedures, the accuracy and reliability of the procedures, and the associated costs and duration. In Sect. 3, we explore three topics that have an impact on the transformative nature of AI regarding accountants and preparations for change and seek to understand how AI affects the role of an accountant and gather opinions on the potential dangers and necessary steps for accountant preparation.

In Sect. 4, demographic data were collected through three questions on age, education level, and years of experience in accounting. In Sect. 5, using three questions, we examine participants’ involvement in AI in accounting, asking about their use of AI programs, knowledge of AI implementation in accounting, and gender. In Sect. 6, using three questions, participants’ awareness and use of AI in accounting covered topics such as awareness of AI advantages, awareness of risks and challenges, and frequency of using AI in accounting procedures. The questionnaire employs a 5ـــpointـــLikert scale to allow participants to express their views on the topics outlined. All the items are included in Appendix 1.

The participants of this study were accounting professionals in Saudi Arabia. Snowball sampling is also performed. A link was created for respondents to view the surveys they received through Google Forms. Google Forms is a simple, straightforward, and user-friendly interface. We personally emailed the survey link to accounting professionals. We also sent the survey link through social media platforms such as Instagram and WhatsApp. We asked accounting professionals' followers to gather data by tagging SOCPA’s account. The social media platforms used for the survey were Instagram, WhatsApp, Facebook, and SnapChat X. The surveys were conducted from September to December 2023. The data were collected during the same period. A total of 202 experts completed the questionnaires.

4 Research findings

The model was analyzed using PLS with the SmartPLS 4 software [80]. As suggested in the literature, a two-step analytical approach is followed. First, the validity and reliability of the measurement model are checked [81]. Subsequently, we tested the hypothesized relationships in the structural model and followed the methodologies suggested in [82]. In addition, the significance of path coefficients and loadings was assessed through the bootstrapping method with 5000 resamples, as suggested by [82].

The sample characteristics illustrated in Table 1 reveal that the dataset was quite diverse and included respondent information from different age groups, gender, educational background, professional experience, and company type. The majority of the respondents consisted of males and females between 18 and 44 years of age. Attitudinal trends regarding AI were generally good, and 89% of the respondents had a positive impact. More training was also needed for AI and related skills, specifically in “more training on AI” (39%). These findings provide a general view of the demographic profile, attitude, and perception of respondents concerning the adoption of AI and the associated implications.

Table 1 Sample demography
Full size table

The descriptive statistics provided in Table 2 offer valuable insights into the distribution characteristics of the various survey items within the dataset and are obtainable in Table 1. The components that constitute these statistics are mean, standardـــdeviation, standardـــerror of the mean, skewness, and kurtosis. Outstandingly, variables such as Q10AI-Rep-Use exhibited a symmetric distribution with a mean and median of 2.347, signaling moderate variability (SD = 0.930), positive excess kurtosis (1.298), and a slight skew to the right (0.885). Conversely, the Q12AI-Eff-ACCT-Prosdu test showed symmetry (mean = median: 3.871), moderate variability (SD = 0.736), high excess kurtosis (2.871), and a pronounced left skew (− 1.135). Similar patterns of symmetry and variability are observed in Q13AI-Effen-Acurat, Q14AI-Enc-Und-Finacil, Q15AI-Enc-Analyses, Q16AI-Enc-AC-Un, Q17AI-Trust, and Q19AI-Accts-Role, each with distinct levels of excess kurtosis and skewness. Additionally, Q1Ege and Q2Gender displayed symmetrical distributions with varying levels of variability, excess kurtosis, and skewness, while Q20AI-Weakness, Q21AI-Traning, Q3Educa, Q4Exper, Q7AI-UnLev, Q8AI-Pors, and Q9AI-Cons provided further insights into the diverse patterns within the dataset. These statistics collectively enhance our understanding of central tendencies, variability, kurtosis, and skewness, facilitating a comprehensive interpretation of the dataset's underlying trends and patterns.

Table 2 Descriptive statistics
Full size table

4.1 Measurement model analysis

4.1.1 Convergent validity

The analysis of the measurement model involved evaluating two types of validity: convergent and discriminant. Convergentــــvalidity was assessed by inspecting the loadings, averageـvarianceــextracted (AVE), and composite reliability following established guidelines (Gholami et al., 2013; Rahman et al., 2015). The loadings, composite reliabilities, and AVEs for all constructs surpassed the recommended thresholds, with loadings exceeding 0.708, composite reliabilities surpassing 0.7, and AVE values exceeding 0.5, as detailed in Table 3 and Fig. 1, aligned with the recommendations of the literature.

Table 3 ConvergentـــValidity
Full size table

4.1.2 Discriminant validity

In evaluating discriminant validity, recent critiques have questioned the reliability of the Fornell and Larcker [83] criterion, leading [84] to propose a multitrait-multimethod matrix approach. Using the heterotrait–monotrait ratio of correlations validated through a Monte Carlo simulation study, we employed this alternative method and found that all values in Table 3 exceeded the recommended thresholds of 0.78 and 089 [85], confirming the establishment of discriminant validity (see Table 4).

Table 4 Discriminantـــvalidity
Full size table

4.2 Structural model analysis

4.2.1 Testing model fit

Based on Table 5 and prior to the testing process, we conducted a model fit evaluation using SRMR, bootstrapped statistical inference, and NFI. SRMR is a measure of the disparity between the observed and model-implied correlation matrices, with values below 0.08 indicating a satisfactory fit. While SRMR was initially developed as a goodness-of-fit measure for partial least squares structural equation modelling (PLS-SEM), it can also be employed to identify model misspecifications. The NFI is an incremental fit measure that compares the chi-square value of the proposed model to a benchmark, with values greater than 0.9 indicating an adequate fit.

Table 5 Testing model fit
Full size table

4.3 Results of the hypothesis testing

The hypothesis testing results for the structural model were assessed based on the recommendations of [82]. The evaluation involved the inspection of R2, beta (β), and the corresponding tــvalues via bootstrapping with 5000 resamples. Additionally, we followed the explicit reporting standards advanced by [89] for quantitative research, which encompass replication studies, effect size estimates, confidence intervals, Bayesian methods, Bayes factors, and decision-theoretic modelling. Therefore, our examination included measures of effect sizes along with their respective confidence intervals, as shown in Table 6 and Fig. 2.

Table 6 Results of the hypothesis testing
Full size table
Fig. 2
figure 2

Bootstrapping results

Full size image

Our investigation focuses on the direct and indirect relationships in the context of AI’s effect of AI on accounting procedures. Significant findings emerged from the analysis of direct relationships. The hypothesis was strongly supported by the way coefficient of 0.311 for H1, standard deviation of 0.092, t-statistic of 3.378, and p-value of 0.001. H2 played a crucial role, as indicated by its coefficient of − 0.650, standard deviation of 0.049, t-statistic of 13.201, and p-value of 0.000. H3 also demonstrated significance, with a regression coefficient of 0.182, standard deviation of 0.100, a t-statistic of 1.823, and p-value of 0.068. Similarly, H4 and H5 were maintained but to differing extents. However, the excess hypotheses (H6 to H9) did not garner much support, based on their independent coefficients, standard deviations, or t-test results. Remarkably, H10, which connects the impact of AI on accounting prosthesis to the impact on accounting proficiency, received substantial support, as evidenced by its 0.833 way coefficient, 33.980 t-statistic, and 0.000 p-value.

H13 had a significant mediation analysis result where standard deviation was 0.038, path coefficient was 0.068, tــstatistic was 1.757, and pــvalue equaling 0.079. On the same note, H14 was also substantial with a path coefficient of 0.152, standard deviation of 0.083, tــstatistic being 1.822, pــvalue equaling 0.068. In the end H15 became significant by having his pــvalue at 0.029 which meant he had a path coefficient of 0.140, Standard Deviation at 0.064 and tــstatistic of 2.190. These findings collectively enhance our understanding of the relationships within the model and shed light on both the direct and mediating effects. H13, H14, and H15 investigate how AI awareness and usage affect various accounting results by engaging in AI and realizing their outcomes. Evidence exists for each of the three hypotheses, affirming that a complex interaction of these variables is essential for an in-depth understanding of AI’s implications of AI in accounting.

Taken together, these results show the complexity of the relationships within a model, where AI also plays many roles in accounting. Thus, this research provides us with useful information concerning both mediate and direct impacts from AI, which enables a better understanding of how AI awareness, engagement, and impact collectively shape the accounting practice landscape..

5 Discussion

In summary, this study investigates how AI is being used or is likely to be used in accounting procedures. We evaluated the levels of comprehension, knowledge, attitudes, and practices regarding the application of AI in accounting procedures among a vast population of accounting professionals. We conclude the literature review based on five core concepts. Under the Fourth Industrial Revolution, many authors have indicated that AI has revolutionized the accounting profession. This transformation is manifested by automating tasks, enhancing precision, and allowing the detection of fraudulent activities, which is in line with the findings of this study. For instance, AI is predicted to quadruple the world’s per capita GDP growth rate by 2035 if it develops at an annual rate of global economic growth. The survey research methodology used a questionnaire as the data collection tool and involved collecting data on demographics, AI knowledge, attitudes about perceptions and practices, and the PLS technique for analysis, where convergent and discriminant validity were emphasized. This study successfully established the validity of the measures, confirming their reliability and distinctiveness.

The governing factor of the t-statistic significance in the structuralــmodel analysis is that most of the hypothesized relationships they test are directly relevant, and thus the t-statistics are very significant, especially when considering the hypotheses concerning the effect of AI on the various dimensions of accounting. The strong support for H1–H5 and H10 appropriately reflects the transformative power of AI. However, the t-statistic supplied to the direct relationships between H6 and H9 fluctuates; hence, a range of t-statistics scores, given that H1 to H5 and H10 demonstrate the highest t-statistics, because the mediator is not a priori set to be an influencer. The mediation analysis (in H13, H14, and H15) of indirect relationships through several mechanisms provides a better understanding of these combinations’ interactions by explaining the various AI chains of effects on accounting outcomes. These findings are also supported by previous studies [e.g. 4, 67, 90–94].

The direct relationships between AI awareness and usage and working with accountants, AI engagement and accountants, and AI and accounting procedures based on the analysis are consistent with the assumptions expressed in the literature review section [e.g4, 5, 95, 96]. This means that accountants who are informed about and using AI are likely to be involved in AI, and consequently, changes in the accounting procedure will be positive. In terms of AI awareness and usage, the path coefficient of 0.311 seems to have a strong positive relationship. This occurs when a higher level of AI awareness aligns more positively with that used in accountants’ work. In contrast, the findings regarding AI awareness and usage were negative, with a very intense level of -0.650. This means that those who are more aware of or actively use AI are similarly likely to utilize it. It appears that they regarded AI as a versatile tool. The path coefficient is 0.833, which indicates a strong positive relationship between “the impact of AI accounting on accounting” and “accounting efficiency.” Therefore, this influence is significant.

For indirect relationships, the analysis found a strong relationship between AI awareness and usage, AI engagement and accountants, and AI awareness and preparation for change. This implies that it is more likely that accountants aware of and using AI are likely to be ready for the variations brought by AI to the accounting profession. AI Awareness and usage had a path coefficient of 0.182, which is a weak positive relationship with preparation for change. AI engagement in accountants has a path coefficient of − 0.055, which is a weak negative relationship with preparation for change. It should be mentioned that although the MAR still had no significant relationships between respondents’ demographic information and experience, AI’s impact on accounting procedures or AI engagement. It can be inferred that this factor did not affect the adoption of AI in the accounting profession.

Moving to the mediation analysis, the current study tests the indirect links of AI awareness and use with several accounting outputs and reveals that AI use and AI impact mediate the relationship between AI awareness and these outputs, which is also in line with prior research [e.g. 67, 90, 97]. There are specifically identified positive and significant relationships in H13, H14, and H15, denoting AI awareness and usage influencing AI engagement and its impact, which positively impacts accounting outcomes. These results, coupled with the evidence that a direct linkage of AI awareness and usage with positive accounting outcomes is consistent, again indicate that the effect is mediated by the level of engagement with AI and its impact. In practical terms, firms investing in raising awareness and the use of AI, along with strategies to enhance engagement and maximize impact, are more likely to result in positive accounting practices. A study like this would imply in a practical sense that firms would invest in AI awareness, use, and engagement strategies to optimize outcomes in accounting. We urge researchers to further investigate the mechanisms through which engagement and the impact of AI mediate the relationships between awareness of AI and usage and the accounting outcome.

In summary, the results are relatively robust in drawing the conclusion that AI has a positive impact on the accounting profession, thereby validating previous evidence on the same matter [e.g. 4, 98]. Accountants who are aware of and use AI will most probably involve and report positive changes in the way accounting procedures are performed. The results show the complexity and interrelated relationships of AI awareness, engagement, and impact, hence giving the broadness of the collective impacts of the future landscape of accounting. In general, this study provides respected insights that may be useful in practice and guide future studies in the dynamic field of accounting.

The current study is in line with the literature on the transformative impact of AI on accounting procedures, especially for those in Saudi Arabia. Therefore, the findings emphasize the economic gains associated with AI, reflecting the McKinsey-Global-Institute's assessment of a 1.2 percent annual rise in global GDP and total economic output of $13 trillion by 2030 [99]. Moreover, economic development rates will increase fourfold globally by 2035, and the impact of AI will be more serious [100]. As the current literature review shows, AI has a twofold influence on accounting functions: elimination of repetitive routines, opportunity to automate data input, and improving fraud detection. The methodology chosen for the current study—questionnaire-based survey and the PLS technique of analysis—is aligned with previous research and hence indicates methodological rigor. The measures are valid based on the findings of discriminately and convergently valid measures. The structural model further explicates that the direct relationships vary in their levels of significance, thereby emphasizing only the nuanced effect that AI has on the dimensions of accounting. This research contributes to how AI affects the accounting profession in a far more granular framework: training and how it affects future economic implications. The context applied in this study is unique to Saudi Arabia. Mediation aids in providing more details on the dynamic interplay between AI exposure and experience and their combined effects on accounting outcomes.

6 Conclusion

The current research thus offers insight with a detailed analysis of how AI impacts accounting procedures within Saudi Arabia. The results strongly confirm reformative influence, both directly and indirectly. This incorporation of AI into accounting processes is expected to bring efficiency, better accuracy, and new challenges that foster continual adaptation and learning. This study provides significant insights for accounting practitioners in Saudi Arabia that may further help develop a deeper understanding of the potential implications of the use of AI in the practice of accounting. In addition, this study builds a good foundation for future research on the dynamic landscapes of AI and its integration into the accounting process.

In other words, with our sharing of applications and implications, we add to the growing body of literature regarding AI in accounting in general, and the role it plays in accounting practitioners in Saudi Arabia in particular. Therefore, such findings are in line with the existing literature; through the addition of nuance on the level of contextual specifics, it enriches the implications of the transformative potential of AI. As AI continues to form part of future economies and industries, the study is a baseline and reference point for organizations and policymakers amidst the increasingly dynamic landscape of AI in accounting. Collective insights from numerous studies underline the need for continuous research and change in response to technological advances.

Therefore, this study proposes several actions that need to be taken by accounting professionals in Saudi Arabia. First, there is a strong recommendation that organizations should integrate AI into accounting practices. To optimize the benefits, there has been a significant highlight of automation, increased accuracy, and improved fraud detection. Organizations also need in-depth training programs, pointing to the need for education in determining the magnitude of the impacts that AI will have on accounting processes. The next important step is to create a continuous adaptation culture in which AI is dynamic and the accounting landscape changes. Strategic planning of economic growth can be conducted in a way that aligns with broader developmental goals through AI adoption by policymakers and organizations. However, the limitations of this study cannot be generalized to various industries or regions. The survey method introduces considerations of response bias; hence, further research is needed to explore alternative data sources. In addition, the dynamic nature of AI makes research equally imperative to be unceasing and adaptable in this field. Moreover, human factors, ethical considerations, and socioeconomic factors need to be more thoroughly investigated to obtain a broader view of the impact AI will have on accounting practices. In addition, a more important problem related to improving the value of accounting information in which research has been conducted is that AI will take on its own control. That is, there is also plays a role in increasing the efficiency of internal control systems in companies and the process of external auditors' auditing of financial statements in Saudi Arabia. Although this study provides valuable insights and recommendations for action, it is necessary to consider these limitations so that organizations and researchers can refine implementation strategies to realize a deep, rich understanding of the multifaceted impact of AI in accounting.