AI Detects 'Signs' of Expense Fraud. Stena Expense Releases New Feature to Automatically Identify High-Fraud-Risk Applicants
ChillStack Inc. has launched a new 'Risk Ranking Function' for its automated expense fraud detection cloud service, Stena Expense. This AI-powered feature analyzes expense data to calculate risk scores for each applicant, visualizing those with high fraud risk in a ranking format. It aims to enable early detection of fraud and streamline audit operations by allowing auditors to prioritize investigations.
📋 Article Processing Timeline
- 📰 Published: April 14, 2026 at 20:00
- 🔍 Collected: April 14, 2026 at 11:31
- 🤖 AI Analyzed: April 14, 2026 at 17:17 (5h 45m after Collected)
ChillStack Inc. (Headquarters: Shibuya-ku, Tokyo; Representative Director: Michiaki Ito; hereinafter 'ChillStack'), which supports government agencies and major corporations in their challenges with AI and security, has launched the 'Risk Ranking Function' as a new feature for its automated expense fraud detection cloud service, 'Stena Expense'.
This function uses AI to analyze expense data, calculate a risk score for each applicant, and visualize high-fraud-risk applicants in a ranking format. This allows auditors to identify individuals who should be prioritized for review, simultaneously achieving early detection of fraud and increased efficiency in audit operations.
Background of Development: Worsening Internal Fraud and 'Delayed Discovery'
Expense fraud is difficult to expose through line-item visual checks, and coupled with the small amount per case, it tends to occur frequently and persist for long periods. According to a survey by the Association of Certified Fraud Examiners (ACFE), the average period until fraud is discovered is 18 months. (*1)
To minimize prolonged damage, rapid and accurate detection technology for early discovery is required. However, there are limits to visually checking vast amounts of expense data within a limited time, and human line-item checks alone make it difficult to spot statistical anomalies. As a result, this contributes to audits becoming a mere formality and fraud becoming normalized.
To solve these challenges and enhance the soundness of expense usage, Stena Expense has newly developed the 'Risk Ranking Function'. This function quantifies applicants' expense usage patterns using multiple risk indicators and visualizes high-fraud-risk applicants in a ranking format. By analyzing statistical usage trends, which were difficult to identify through line-item visual checks, and automating the prioritization of audit targets, it simultaneously achieves early detection of fraud and advanced auditing.
(*1) Source: Association of Certified Fraud Examiners (ACFE) 'Occupational Fraud 2024: A Report to the Nations' https://www.acfe.jp/wp-content/uploads/2024/07/RTTN_2024_JP1-.pdf
About 'Stena Expense's' Risk Ranking Function
'Stena Expense' is a system that automatically detects fraud, anomalies, and inappropriate expense items based on corporate employee expense claim and approval data. The 'Risk Ranking Function' analyzes vast amounts of expense data using its unique AI algorithm to calculate a risk score for each applicant based on past fraud patterns and unnatural application trends. By visualizing applicants with high fraud risk in a ranking format based on this score, it allows for the identification of individuals who should be prioritized for review.
Key Features
Multi-faceted Risk Assessment via Radar Chart
Applicants' expense usage patterns are evaluated across '9 risk indicators' and visualized in a radar chart. By comprehensively understanding application trends within a department, patterns of fraud that might be overlooked by mere numbers can be identified at a glance.
Presentation of Countermeasures Based on Judgment Results
Based on the calculated 9 risk indicators and detection results, a risk assessment result is generated for each applicant. Three alert levels (high risk, caution, information) are assigned according to the nature of the risk, and key points to check and specific countermeasures (e.g., cross-referencing with original receipts, detailed examination of transactions with specific business partners) are presented. This allows auditors not only to identify items to check but also to smoothly determine initial response actions.
Webinar Announcement
To commemorate the release of this function, we will hold a panel discussion featuring Mr. Saito, CEO of Rick & Company, who has supported internal controls in numerous companies.
Internal Control Update in the DX Era | Expense Management for 'Offense and Defense'
Date & Time: Monday, April 20, 2026, 11:00 AM - 12:00 PM JST
Format: Online (Zoom Webinar)
Participation Fee: Free (Pre-registration required)
Speakers: Yoshiaki Saito (Certified Public Accountant, Tax Accountant), Representative CEO, Rick & Company LLC
Hiroyuki Nakamichi, Director & COO, ChillStack Inc.
Apply for the webinar
About ChillStack Inc.
Company Overview
Company Name: ChillStack Inc.
Location: Yoyogi Air Heights 206, 5-16-10 Sendagaya, Shibuya-ku, Tokyo
Founded: November 2018
Representative Director: Michiaki Ito
Corporate Website: https://chillstack.com/
Business Activities
'Supporting societal innovation with cutting-edge AI and security technology.'
ChillStack provides solutions based on world-class AI security technology that comprehensively resolve risks associated with the development of AI and DX. For corporations, we offer services that support fraud/anomaly analysis and secure AI utilization. We also collaborate with government agencies to promote research and development and social implementation for solving more complex and advanced societal issues.
Awards: EY Innovative Startup 2026
*Partial Excerpt
・Development and provision of 'Stena Expense', an AI system that automatically inspects expense fraud and deficiencies.
https://expense.stena.chillstack.com
・Service 'Stena AI' that promotes generative AI utilization through automated detection of fraudulent use.
https://ai.stena.chillstack.com
・Development and provision of 'Security Diagnosis' to identify security risks in services.
https://pentest.chillstack.com
・Research and development on AI security measures, education business 'AI Defense Institute'.
https://jpsec.ai
Representative Director & CEO Profile
Michiaki Ito
Engaged in AI x security research, awarded the Best Paper Award at the international conference IEEE CSPA2018 and the IPA Security Camp Award 2018 Grand Prize.
He has won international security competitions and is involved in nurturing the next generation of AI security talent, serving as an instructor for AI security lectures at Security Camp 2019 - 2025. Selected for 'Forbes 30 Under 30 Asia 2025'.
Recruitment Information
ChillStack is currently looking for colleagues who resonate with our mission & vision and are willing to take on challenges together. We are always open for information exchange and casual interviews for those interested in learning more about ChillStack, so please feel free to contact us. We look forward to your application.
https://chillstack.com/career
Note
ChillStack's note features employee interviews and the untold story of 'Stena Expense's' birth. Please take a look.
https://note.com/chillstack