Head of the Research Group Engineering Mathema-cs and Compu-ng Lab (EMCL) Interdisciplinary Center for Scien-fic Compu-ng (IWR) Heidelberg University • Head of the Research Group Data Mining and Uncertainty Quan-fica-on (DMQ) Heidelberg Ins-tute for Theore-cal Studies (HITS gGmbH)

Chief Informa-on Officer (CIO) Heidelberg University, Germany  Spokesman IT Security for all Universities State Baden-Württemberg, Germany  Director of the Heidelberg University Compu-ng Centre (URZ) Reference: From Rules to Learning: Is AI Transforming or Just Enhancing IDS? The rapid advancement of Artificial Intelligence (AI) has profoundly impacted cybersecurity, particularly in the domain of Intrusion Detection Systems (IDS). While traditional IDS rely on rulebased and signature-based methods, AI introduces adaptive learning, anomaly detection, and predictive analytics—raising a critical question: Is AI an evolution or a revolution for IDS? This talk explores whether AI represents a natural progression (evolution) of existing IDS techniques or a disruptive shift (revolution) that fundamentally transforms threat detection. We examine key AI-driven advancements, such as machine learning for behavioral analysis and deep learning for zero-day attack detection, while addressing challenges like adversarial attacks, explainability, and data dependency. By analyzing real-world implementations and emerging trends, we assess whether AI augments traditional IDS frameworks or necessitates a complete paradigm shift. This talk aims to foster the understanding of AI’s role in IDS, helping stakeholders prepare for both its opportunities and risks.