Research project 2026 Completed
Research on adversarial content rewriting, prompt injection and ensemble defences for LLM-powered social media bot detection.
Research question
Nof Orenstein and Yoni Birman study attacks and defences for language-model systems that identify social media bots. The work examines adversarial content rewriting and prompt injection, and introduces LSABRE, an ensemble approach to improving robustness.
The linked outputs include an IEEE Symposium on Security and Privacy 2026 poster and a journal article in Pragmatic Cybersecurity. The journal article extends a presentation at ACISP 2026.
Pragmatic Cybersecurity, 2026, 1(2), Article 10
A study of attacks and defences for language-model systems that identify social media bots.
IEEE Symposium on Security and Privacy 2026 · Poster
An IEEE Symposium on Security and Privacy 2026 poster on adversarial attacks and ensemble defences for LLM-based bot detection.
Explore the people and research areas connected to this project.
Discuss collaboration ↗
