<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE ArticleSet PUBLIC "-//NLM//DTD PubMed 2.7//EN" "https://dtd.nlm.nih.gov/ncbi/pubmed/in/PubMed.dtd">
<ArticleSet>
<Article>
<Journal>
				<PublisherName>Univrsity Of Tehran Press</PublisherName>
				<JournalTitle>Accounting and Auditing Review</JournalTitle>
				<Issn>2645-8020</Issn>
				<Volume>33</Volume>
				<Issue>3</Issue>
				<PubDate PubStatus="epublish">
					<Year>2026</Year>
					<Month>09</Month>
					<Day>23</Day>
				</PubDate>
			</Journal>
<ArticleTitle>Identifying and Prioritizing the Key Components of Artificial Intelligence Applications for Enhancing Independent Auditing Processes and Mitigating Financial Risks: A Fuzzy Delphi and Thematic Analysis Approach</ArticleTitle>
<VernacularTitle>Identifying and Prioritizing the Key Components of Artificial Intelligence Applications for Enhancing Independent Auditing Processes and Mitigating Financial Risks: A Fuzzy Delphi and Thematic Analysis Approach</VernacularTitle>
			<FirstPage>481</FirstPage>
			<LastPage>521</LastPage>
			<ELocationID EIdType="pii">108615</ELocationID>
			
<ELocationID EIdType="doi">10.22059/acctgrev.2026.410148.1009236</ELocationID>
			
			<Language>FA</Language>
<AuthorList>
<Author>
					<FirstName>Hasan</FirstName>
					<LastName>Hamed Mohammad</LastName>
<Affiliation>PhD. Candidate, Department of Accounting and Financial Management, Faculty of Economics and Management, Urmia University, Urmia, Iran.</Affiliation>
<Identifier Source="ORCID">0009-0006-2669-1788</Identifier>

</Author>
<Author>
					<FirstName>Mehdi</FirstName>
					<LastName>Heidari</LastName>
<Affiliation>PhD. Candidate, Department of Accounting and Financial Management, Faculty of Economics and Management, Urmia University, Urmia, Iran.</Affiliation>
<Identifier Source="ORCID">0000-0002-0148-1470</Identifier>

</Author>
<Author>
					<FirstName>Ali</FirstName>
					<LastName>Ashtab</LastName>
<Affiliation>Associate Prof., Department of Accounting and Financial Management, Faculty of Economics and Management, Urmia University, Urmia, Iran.</Affiliation>
<Identifier Source="ORCID">0000-0002-7339-2100</Identifier>

</Author>
</AuthorList>
				<PublicationType>Journal Article</PublicationType>
			<History>
				<PubDate PubStatus="received">
					<Year>2026</Year>
					<Month>01</Month>
					<Day>27</Day>
				</PubDate>
			</History>
		    <Abstract>Objective
The primary objective of this research is to design and validate a structural model for the application of Artificial Intelligence (AI) aimed at mitigating financial risks within the context of independent auditing.
Methods
This study adopts an applied-developmental design, employing a mixed-methods (qualitative-quantitative) approach. In the qualitative phase, following a systematic literature review and the filtration of  initial sources down to  core references, an inductive thematic analysis was conducted alongside semi-structured interviews with  subject-matter experts. The expert panel—comprising university professors specializing in AI, senior managers and partners of audit firms, and IT/AI specialists was selected utilizing purposive and snowball sampling techniques. This phase yielded  open codes, which were subsequently classified into  overarching categories: predicting financial trends, task automation, financial data analysis, fraud detection, enhancing internal controls, increasing accuracy and efficiency, time optimization, improving audit quality, timely auditing, and workforce reduction. The inter-rater reliability of the coding process was corroborated, yielding a Kappa coefficient of  . In the quantitative phase, the proposed model was validated by an independent panel of  experts via the Fuzzy Delphi method (utilizing a  -point scale, with an acceptance threshold of  ). Furthermore, the causal relationships and the hierarchical structure of the identified components were analyzed applying Decision-Making Trial and Evaluation Laboratory (DEMATEL) and Interpretive Structural Modeling (ISM) techniques.
Results
The comparative ranking of the indicators demonstrated that timely auditing acquired the highest priority (mean =  ), followed by increasing accuracy and efficiency (mean =  ), and improving audit quality (mean =  ). The DEMATEL analysis revealed that fraud detection and error elimination constitute the most influential components, functioning as central catalysts for the enhancement of audit quality. Additionally, task automation exhibited a substantial causal impact on time optimization (coefficient =  ) and the prediction of financial trends (coefficient =  ). Conversely, workforce reduction emerged as the least impactful factor, underscoring the complementary nature of AI relative to human cognitive functions. 
Conclusion
The findings conclude that the integration of AI into independent auditing predominantly expedites the audit process while significantly augmenting its accuracy, efficiency, and overarching quality. The fundamental efficacy of AI is rooted in its capacity for fraud and anomaly detection coupled with process automation; thereby rendering workforce reduction a marginal and secondary consequence. The originality of this research stems from the synergistic application of thematic analysis and the Fuzzy Delphi method to develop a localized AI adoption model tailored to the Iranian independent auditing environment. By leveraging a rigorous mixed-methods design and drawing upon expert consensus, this study introduces a pragmatic framework conducive to professional policymaking, auditor upskilling, and future academic inquiries into the digital transformation of the auditing profession. Ultimately, it contributes to the extant literature by delineating a robust mechanism for mitigating financial risks through AI utilization.</Abstract>
			<OtherAbstract Language="FA">Objective
The primary objective of this research is to design and validate a structural model for the application of Artificial Intelligence (AI) aimed at mitigating financial risks within the context of independent auditing.
Methods
This study adopts an applied-developmental design, employing a mixed-methods (qualitative-quantitative) approach. In the qualitative phase, following a systematic literature review and the filtration of  initial sources down to  core references, an inductive thematic analysis was conducted alongside semi-structured interviews with  subject-matter experts. The expert panel—comprising university professors specializing in AI, senior managers and partners of audit firms, and IT/AI specialists was selected utilizing purposive and snowball sampling techniques. This phase yielded  open codes, which were subsequently classified into  overarching categories: predicting financial trends, task automation, financial data analysis, fraud detection, enhancing internal controls, increasing accuracy and efficiency, time optimization, improving audit quality, timely auditing, and workforce reduction. The inter-rater reliability of the coding process was corroborated, yielding a Kappa coefficient of  . In the quantitative phase, the proposed model was validated by an independent panel of  experts via the Fuzzy Delphi method (utilizing a  -point scale, with an acceptance threshold of  ). Furthermore, the causal relationships and the hierarchical structure of the identified components were analyzed applying Decision-Making Trial and Evaluation Laboratory (DEMATEL) and Interpretive Structural Modeling (ISM) techniques.
Results
The comparative ranking of the indicators demonstrated that timely auditing acquired the highest priority (mean =  ), followed by increasing accuracy and efficiency (mean =  ), and improving audit quality (mean =  ). The DEMATEL analysis revealed that fraud detection and error elimination constitute the most influential components, functioning as central catalysts for the enhancement of audit quality. Additionally, task automation exhibited a substantial causal impact on time optimization (coefficient =  ) and the prediction of financial trends (coefficient =  ). Conversely, workforce reduction emerged as the least impactful factor, underscoring the complementary nature of AI relative to human cognitive functions. 
Conclusion
The findings conclude that the integration of AI into independent auditing predominantly expedites the audit process while significantly augmenting its accuracy, efficiency, and overarching quality. The fundamental efficacy of AI is rooted in its capacity for fraud and anomaly detection coupled with process automation; thereby rendering workforce reduction a marginal and secondary consequence. The originality of this research stems from the synergistic application of thematic analysis and the Fuzzy Delphi method to develop a localized AI adoption model tailored to the Iranian independent auditing environment. By leveraging a rigorous mixed-methods design and drawing upon expert consensus, this study introduces a pragmatic framework conducive to professional policymaking, auditor upskilling, and future academic inquiries into the digital transformation of the auditing profession. Ultimately, it contributes to the extant literature by delineating a robust mechanism for mitigating financial risks through AI utilization.</OtherAbstract>
		<ObjectList>
			<Object Type="keyword">
			<Param Name="value">Audit process improvement</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Content analysis</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Independent Auditing</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Financial Risks</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Artificial Intelligence</Param>
			</Object>
		</ObjectList>
<ArchiveCopySource DocType="pdf">https://acctgrev.ut.ac.ir/article_108615_c17d0a405ecd22f0969717ae8143d9c7.pdf</ArchiveCopySource>
</Article>
</ArticleSet>
