Eyes on the Table An AI-Powered Casino Fraud Detection System That Monitors Every Chip Exchange

A Japanese gaming technology company has patented a sophisticated casino surveillance system that uses artificial intelligence and image analysis to automatically detect fraudulent chip exchanges, cash swaps and settlement errors at gaming tables. The system monitors every game in real time, flagging suspicious transactions the moment they occur – giving casino operators a powerful tool to protect the integrity of their gaming floors.

The Problem

Casino fraud is a persistent and costly challenge. Despite the heavy investment casinos make in surveillance equipment and trained security staff, sophisticated cheating methods – particularly those involving the manipulation of chips and cash at the table – remain difficult to detect reliably in real time. Fraudsters are well aware of the limitations of human observation and traditional camera monitoring, and they exploit moments of distraction, fast dealing sequences and the complexity of multi-player table games to carry out their schemes.

The specific problem targeted by this invention is the detection of unauthorised cash-for-chip exchanges at gaming tables. In a legitimate casino operation, cash can only be exchanged for chips under defined conditions, and settlements (paying out winnings or collecting losing bets) must be carried out in accordance with game rules. When these transactions happen outside the correct sequence – for example, when a player or a dishonest dealer exchanges cash for chips while the dealer is not in the process of settlement – it may signal fraud or error.

Traditional surveillance cameras generate enormous volumes of footage that security staff cannot monitor comprehensively in real time. Rule-based detection systems struggle with the variability of real game play. What is needed is an intelligent system that understands the game state and can automatically distinguish legitimate transactions from suspicious ones.

What This Invention Does

Angel Group’s fraud detection system connects multiple gaming tables into a unified monitoring architecture. Each table is equipped with a game recording apparatus (camera), a card distribution device (which determines game outcomes) and an image-analysing apparatus. A central control device ties these elements together.

The card distribution device provides the system with real-time knowledge of the current game state – specifically, whether or not the dealer is currently in a dealing phase. The image-analysing apparatus uses AI to monitor the table and detect the presence and movement of cash and chips. Crucially, the system incorporates an artificial intelligence model trained to identify cash-for-chip exchanges that occur outside of legitimate settlement conditions – for example, when the dealer is not engaged in settlement according to the card distribution device’s records.

When such an anomaly is detected, the system can flag the event automatically, enabling rapid response by security personnel. By combining game-state knowledge with AI-driven visual analysis, the system avoids many of the false-positive issues that plague simpler detection approaches.

Key Features

Game-state-aware fraud detection. By linking the AI detection model to real-time information from the card distribution device, the system understands when chip and cash exchanges are and are not appropriate – enabling far more accurate fraud detection than camera-only systems.

AI-powered image analysis. An artificial intelligence model trained to detect cash and chip presence and movement provides automated, continuous monitoring across all tables, independent of human attention or reaction time.

Multi-table centralised monitoring. The architecture supports a plurality of gaming tables feeding into a single control device, enabling casino-wide monitoring from a centralised point and facilitating coordinated responses to detected anomalies.

Real-time detection capability. Events are flagged as they happen, rather than being identified in a post-game review – giving security teams the opportunity to intervene before a fraudulent transaction is completed.

Reduced false positives through context awareness. The integration of game-state data means the system can distinguish normal game activity from suspicious behaviour, reducing the rate of false alarms that can desensitise security teams.

Who Is Behind It?

Angel Group Co., Ltd. is a Japanese gaming technology company with a portfolio of innovations spanning casino surveillance and gaming systems. The application traces its priority to Japanese Patent Applications filed in August and October 2015, reflecting a decade of development in this space. This Australian application is a divisional of AU 2025204131. The sole inventor is Yasushi Shigeta. The application is managed by Pizzeys Patent and Trade Mark Attorneys Pty Ltd in Canberra.

Why It Matters

Casino fraud costs the global gaming industry hundreds of millions of dollars annually and undermines the regulatory frameworks that licensed operators depend on to maintain their operating licences. For casino operators, the ability to automatically detect fraudulent chip and cash handling in real time represents a significant advance in gaming floor security.

The system is also relevant to regulatory compliance. Casinos are typically required by gaming regulators to maintain comprehensive records of all transactions and to demonstrate the integrity of their operations. An automated, AI-driven detection system that provides a continuous, tamper-resistant record of table activity can strengthen a casino’s compliance posture. With IPC classifications covering casino monitoring (G07C 11/00), casino furniture (A47B 25/00) and computer vision (G06T 7/00), the patent encompasses both the hardware and the intelligence layer of the system.

---

AU 2026201516 was published in the Australian Official Journal of Patents on 19 March 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.

Related Concepts

Casino fraud – particularly chip and cash manipulation at gaming tables – has long challenged surveillance teams reliant on manual monitoring. Artificial intelligence-powered image analysis enables real-time detection of anomalous transactions by combining game-state awareness with continuous visual monitoring. Gaming regulators increasingly require operators to demonstrate robust transaction integrity controls, making automated fraud detection systems valuable for both security and compliance.