EUVIMEDEUVIMED entdecken
Scoping Review

EMERGING FORENSIC ACCOUNTING TECHNOLOGIES FOR FRAUD DETECTION: A COMPREHENSIVE SCOPING REVIEW

Ros Azilatul Nadiah Md Isa, Mohd Danial Afiq Khamar Tazilah

Advanced International Journal of Business Entrepreneurship and SMEs · 30.06.2026 · 8 (28) · 776

DOI: 10.35631/aijbes.828049 ↗

Kurz gefasst

Scoping Review in der Zeitschrift Advanced International Journal of Business Entrepreneurship and SMEs, online erschienen am 30.06.2026. Verfasst von Ros Azilatul Nadiah Md Isa und 1 weiteren Autor:innen. Die Arbeit nennt 55 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As the world becomes digital, financial fraud too is evolving and is a challenge for organisations today to detect and prevent frauds. The fraud schemes get more involved and rely on the latest technology, and traditional audit methods are no longer able to catch up. This scoping review follows the PRISMA-ScR guidelines and the Arksey and O'Malley (2005) framework to analyse how forensic accounting is evolving with the development of new technologies, using a combination of Web of Science and Scopus for literature collection, which results in 116 empirical studies between 2022 and 2024 (89 from Web of Science and 27 from Scopus). Five themes were identified from the analysis. The technology-based fraud detection using artificial intelligence and machine learning was the most notable which was seen in 67 studies (57.8%) with a range of accuracy between 93% and 98.58%. Although the role of technology is becoming increasingly important, there were still 18 studies (15.5%) that emphasized the importance of the auditor's role in supporting the technological tools. Governance and internal controls were found in 21 studies (18.1%) and professional skepticism in 12 studies (10.3%) were organizational and individual factors, respectively. Specifically, the highest accuracy was obtained by using Random Forest (98.58%), K-Nearest Neighbor (95.89%) and Convolutional Neural Networks (95.8%). The hybrid combinations of Multiple Benford's Law models and the K-Means clustering technique were found to be useful for the accuracy of the detection increased from 40% to 93.33%. In terms of geographical distribution, Asian markets are the most prolific and more than 68 studies (58.6%) were published in Asia of which more than 24 (24.1%) were published in China and 12.9% in Indonesia, suggesting the region's high rate of digitalization and dire need for fraud detection. African and Latin American research was very little (4.3% combined) and therefore there was a huge research gap. There were a couple of limitations identified: Most studies (94.8%) were of cross-sectional designs which overlooked longitudinal perspectives and research into implementation challenges is still limited, as is research into the cost benefit analysis and the ethical implications of AI-based fraud detection. This review provides a comprehensive, up-to-date literature review and integrates the latest developments in forensic accounting technology, providing recommendations for organizations, policy makers and researchers.

Zitieren

APA

Isa, R. A. N. M., & Tazilah, M. D. A. K. (2026). EMERGING FORENSIC ACCOUNTING TECHNOLOGIES FOR FRAUD DETECTION: A COMPREHENSIVE SCOPING REVIEW. Advanced International Journal of Business Entrepreneurship and SMEs, 8(28), 776. https://doi.org/10.35631/aijbes.828049

Vancouver

Isa RANM, Tazilah MDAK. EMERGING FORENSIC ACCOUNTING TECHNOLOGIES FOR FRAUD DETECTION: A COMPREHENSIVE SCOPING REVIEW. Advanced International Journal of Business Entrepreneurship and SMEs. 2026;8(28):776. doi:10.35631/aijbes.828049

BibTeX
@article{Isa2026,
  author = {Isa, Ros Azilatul Nadiah Md and Tazilah, Mohd Danial Afiq Khamar},
  title = {EMERGING FORENSIC ACCOUNTING TECHNOLOGIES FOR FRAUD DETECTION: A COMPREHENSIVE SCOPING REVIEW},
  journal = {Advanced International Journal of Business Entrepreneurship and SMEs},
  year = {2026},
  volume = {8},
  number = {28},
  pages = {776},
  doi = {10.35631/aijbes.828049}
}
Kostenlos starten →