AI in Cybersecurity: the Good, the Bad, and the Ugly?
Artificial intelligence has entered cybersecurity with remarkable impact. It detects anomalies in vast datasets, filters endless logs, and automates repetitive tasks. That is the good.
The bad: Attackers adapt just as quickly. Slight modifications can deceive machine-learning classifiers, and LLMs craft phishing messages that look genuine. Defenders have seen AI systems fail spectacularly, generating floods of false alarms or overlooking new attack patterns.
And the ugly? The belief that AI could replace human expertise. It cannot. We need specialists who understand what AI cannot explain, who employ these tools securely, and who judge its benefits and limitations critically.
Hans Peter Reiser
Hans P. Reiser is an Associate Professor at the computer science department at Reykjavik University and co-founder of the Frostbyte cybersecurity research lab. His work explores how to make distributed systems more secure and reliable, bridging the gap between academic research and real-world challenges in cybersecurity.
Hans earned his degree in Computer Science from the University of Erlangen–Nuremberg and his PhD from Ulm University. Before joining Reykjavik University, he held faculty positions at the University of Lisbon in Portugal and the University of Passau in Germany. He has also spent time abroad as a Visiting Professor at Carnegie Mellon University, USA and as a Visiting Full Professor at Karlsruhe Institute of Technology, Germany.
