The AI-ARENA project focuses on developing and evaluating a methodological framework for applying artificial intelligence (AI)-based agents in multidisciplinary research within the digital society. The core scientific approach will be Design Science Research (DSR), within which an AI framework will first be designed as an artifact, followed by a series of case studies in key domains. Each case study includes the development and evaluation of specific AI agents in different contexts: biometrics, digital forensics, information security, computer games, education, queueing theory, social choice theory, and the ethics and bias of AI systems. The agents will serve as tools for analysis, automation, and experimentation within specific problems, and the project will also address the technical, ethical, and regulatory challenges of their use. The goal is to create a flexible and applicable framework that connects AI technology with domain knowledge and scientific methodology, delivering concrete scientific and educational outcomes and contributing to open and responsible science.