Sahar Abdelnabi

Sahar Abdelnabi is a Principal Investigator at the ELLIS Institute Tübingen and an Independent Research Group Leader at the Max Planck Institute for Intelligent Systems. She leads the COMPASS research group (COoperative Machine intelligence for People-Aligned Safe Systems). Her research focuses on AI security, safety, and alignment, with particular expertise in multi-agent systems, prompt injection attacks, privacy frameworks, and evaluation robustness. Prior to her current role, she worked at the Microsoft Security Response Center on AI security vulnerabilities and red-teaming. Sahar’s contributions include pioneering work on indirect prompt injection in LLM-integrated applications, which has been widely adopted by NIST, MITRE, OWASP, and Microsoft. She holds a PhD from CISPA Helmholtz Center for Information Security and Saarland university. Her work received “Best Paper” awards at AISec 2023 workshop and ACL 2025.

AI2050 Project

AI agents will act on our behalf: scheduling appointments, negotiating contracts, coordinating healthcare. For this to work, agents must share information and build on each other’s outputs, relying on existing trust norms and establishing new ones. But trust and unbounded cooperation create vulnerabilities that an attacker may exploit. Abdelnabi’s project studies that tension. They investigate how agents should manage sensitive information across different relationships and how cooperation mechanisms can be subverted through collusion or manipulation. They will design layered defenses, combining architectural safeguards, learned reasoning, and policy frameworks, that keep multi-agent systems both cooperative and secure, ensuring AI serves people safely.