AI in Security and Privacy

Contact person: Nils Gruschka
Keywords: Security AI; MLPrivacy    
Research group: Digital Security (SEC)
Department of Informatics

Like in most disciplines, artificial intelligence is becoming increasingly crucial in Cyber Security and Privacy/Data protection. On one hand, AI allows more efficient and automatic detection and defense against cyber-attacks. On the other hand, AI enables powerful tools for attackers to find weaknesses in IT systems and perform automatic cyber-attacks. An example is social-engineering attacks using deep fake audio and video. In addition, AI can pose severe threats to the privacy of individuals. First, many AI models are trained on personal data, and many attacks exist to recreate this original data. Further, AI allows detailed profiling, and AI models make more decisions (e.g., loan requests).

Research topics:

  • AI-assisted cyber-attack detection
  • Automatic pen-testing/ethical hacking
  • Explainable AI
  • Federated learning
  • Deep fake detection

Mentoring and internship will be offered by a relevant external partner.