common Co., Ltd. Launches 'Kaiji Check AI' to Support Fraud and Risk Analysis of Disclosure Documents

common Co., Ltd. has launched an AI service called 'Kaiji Check AI' (https://www.commoncoltd.com/kaijicheck-ai/), which analyzes disclosure documents (PDFs) from multiple angles to detect signs of fraud and risk, and provides auxiliary information. The service supports U.S. annual reports (Form 10-K), Japanese securities reports, and cryptocurrency whitepapers, evaluating risks from multiple perspectives including rule-based systems and generative AI. Users can instantly verify 'why this conclusion was reached (the rationale)' and 'where it is stated in the document (the relevant section)' directly within the PDF.

It should be noted that this service provides reference information to assist in risk analysis and does not determine or guarantee the presence or absence of fraud. Kaiji Check AI: https://www.commoncoltd.com/kaijicheck-ai/

Kaiji Check AI – Visualizing Fraud Risks: https://www.commoncoltd.com/kaijicheck-ai/

Challenges and Background of Kaiji Check AI Development

Fraud exploiting increased investment interest is on the rise

Disclosure documents are lengthy and require significant time and effort to review

Signs of financial misrepresentation, fraud, and risks are easily overlooked without specialized knowledge

Document formats and analytical focus differ by country and asset type (U.S., Japan, cryptocurrency)

Matching 'rationales' with 'relevant sections' takes time, leading to subjective, person-dependent judgments

Service Overview

Name: Kaiji Check AI

Category: AI service for assisting risk analysis (judgment support) of disclosure documents

Target Documents: U.S. annual reports (Form 10-K), Japanese securities reports, cryptocurrency whitepapers

Target Users: Professionals in investment, credit assessment, research, IR, and strategic planning who review disclosure documents

Delivery Model: SaaS (browser-based, supports Japanese PDFs)

Pricing: Free tier (1 PDF per month), Paid plan: ¥940/month (tax included, 50 PDFs per month)

Key Features

Multi-perspective risk analysis: Evaluates risks using multiple approaches such as rule-based systems and generative AI, providing risk indicators (statistical methods also used for U.S. annual reports)

Visualization of rationale: Clearly explains 'why' a particular judgment was made

Jump to relevant sections: Highlights source statements in the PDF for immediate verification

Supports three jurisdictions/assets: Analyzes documents from Japan, the U.S., and cryptocurrency projects from their respective viewpoints

Practical design considerations: Supports Japanese PDFs, requires login, and uses encrypted communication

Positioning of the Service

The analysis results provided by this service do not determine or guarantee the presence or absence of fraud, misrepresentation, or deception. They are reference materials intended to support human judgment. The service does not provide investment advice, investment solicitation, or professional guidance related to auditing, accounting, or legal matters. Final decisions must be made by users themselves, based on their own responsibility, including verification of original documents and consultation with experts.

[Executive Comment]

"Reviewing disclosure documents is a critical task that requires specialized knowledge and time. Kaiji Check AI aims to streamline this process by quickly highlighting risk indicators and answering 'why' and 'where,' thereby supporting human decision-making. We aim for this tool to serve as a practical reference to assist professionals, strictly as a supplementary aid." (Takeshi Ishikawa, Representative Director)