Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1757
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-17T22:03:17.783671Z
status: published
lens: g_space
sector: crime
headline: Artificial Intelligence Techniques for Financial Fraud Detection: A Bibliometric Analysis of Scientific Development and Future Directions
dek: This study examines the development and intellectual structure of artificial intelligence (AI)-based financial fraud detection through a bibliometric analysis of 2200 Scopus- and Web of Science-indexed journal articles published between 1992 and 2026. The literature has expanded rapidly, with 89.6% of publications appearing since 2018 and an annual growth rate of 29.4%. Research is concentrated in Computer Science and Engineering, whereas Business and Economics account for 9.2% of classified publications. Credit…
gain_title: A review of 2200 indexed articles shows AI fraud-detection research expanding quickly and moving toward deep learning, ensemble methods, graph neural networks and explainable AI.
problem_title: (none)
trace_subject: (none)
gain_reading: A review of 2200 indexed articles shows AI fraud-detection research expanding quickly and moving toward deep learning, ensemble methods, graph neural networks and explainable AI.
gain_evidence: shifted from traditional data-mining and classical machine-learning approaches toward deep learning, ensemble methods, graph neural networks, and explainable AI
problem_reading: (none)
problem_evidence: (none)
quick_read: The study analyzed 2200 journal articles indexed in Scopus and Web of Science from 1992 to 2026. It found that most of the work appeared after 2018, with an annual growth rate of 29.4%, and that credit card and payment fraud is the largest research domain.

The mapping shows where scholarly attention on AI fraud detection is concentrated and which methods dominate. It describes the research literature, not how well these systems detect fraud in practice, so real-world effectiveness is not established by the text.
limitation: 
tag: Evidence-backed gain
key_points: The dataset covers 2200 Scopus- and Web of Science-indexed articles published between 1992 and 2026. | Credit card and payment fraud accounts for 32.0% of research, ahead of banking, loan and transaction fraud at 14.3%. | Insurance fraud makes up 12.2% and accounting and financial statement fraud 9.9% of the literature. | Neural networks and deep learning, ensemble and boosting methods, and classical supervised classifiers are the dominant method families. | Of authors, 88.9% contributed only a single publication.
rundown: Research is concentrated in Computer Science and Engineering, while Business and Economics account for 9.2% of classified publications. This suggests the field is driven mainly by technical disciplines.

Science mapping shows the intellectual structure remains strongly centered on payment fraud, while accounting and insurance fraud form more distinct research fronts. International collaboration varies considerably across leading countries.

The authors combined performance analysis, science mapping, fraud-domain classification, bibliometric laws and publication-growth analysis to assess the field's development and emerging directions.
sources:
- peer_reviewed | Journal of Risk and Financial Management | https://doi.org/10.3390/jrfm19100819 | 2026-10-16
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

87a86e4c73f77e4a28e38154a49b8b103c1db6b2ff82945809dedda3d4d892f1

previous

0000000000000000000000000000000000000000000000000000000000000000