People, Agents, and Security

We study security at the intersection of people, software, and AI agents. Our work examines how people interact with security mechanisms, including phishing prevention, authentication, user interfaces, and secure decision-making. We also investigate risks introduced when agents use programs on people’s behalf, including computer-use, browser, and coding agents. We design and evaluate safer agentic systems and sandboxes, and explore how LLMs can help users make security decisions.

Active Members

Dr. Kari Kostiainen
Senior Scientist
Friederike Groschupp
Doctoral Student
Claudio Anliker
Doctoral Student
Nicolas Dutly
Doctoral Student

Talks

Phishing in Organizations: Findings from a Large-Scale and Long-Term Study

Publication

Selected Publications

Daniele Lain, Tarek Jost, Siniša Matetić, Kari Kostiainen, and Srdjan Čapkun
ACM Conference on Computer and Communications Security (CCS '24)
Distinguished Paper Award