Inference Attacks and Defenses on LLM-Powered Email Phishing Detection Systems

Frederik Sherman and Yoni Birman explore attacks and defenses involving LLM-powered phishing detection, presented at the USENIX Security 2026 poster session.

Frederik Sherman, Dr. Yoni Birman

USENIX Security 2026 · Poster session

Publication

Presented in the USENIX Security 2026 poster session on 12 August 2026.

The research examines inference attacks and defenses for systems that use large language models to identify phishing emails.

Official poster listing

A related paper is available in the HumSec 2026 record.

The next question could be yours

Good research starts
with a conversation.

Students, researchers, collaborators, and partners: explore what we could do together.

Find your way into the lab ↗

Good research startswith a conversation.

Good research starts
with a conversation.