DOUSE : detecting and operating upon instances of self‐exclusion within the iGaming industry

Abstract

The gambling industry has experienced significant global growth in recent years, leading to increased accessibility and participation in gambling activities. While this expansion has created new entertainment opportunities, it has also raised concerns about problematic gambling behaviours that can negatively impact players’ lives. To address these issues, the iGaming industry has begun leveraging artificial intelligence (AI) to detect early signs of problematic gambling and enable timely interventions. This paper presents DOUSE (Detecting and Operating Upon instances of Self‐Exclusion), a system that integrates machine learning models with explainable AI frameworks and large language models to identify players at risk of self‐exclusion. The key components of DOUSE include: • A machine learning model that predicts the likelihood of a player self‐excluding based on their gambling activity patterns • An explainable AI component that provides transparency into the model’s decision‐making process • Integration with a large language model to generate human‐readable explanations of predictions This research aims to contribute to advancing responsible gaming practices in the iGaming industry. The proposed system offers a framework for early detection of problematic gambling behaviours, potentially enabling more timely and effective player protection interventions. By combining predictive power with explainability, DOUSE represents a step towards creating safer and more transparent online gambling environments