Paul-Danut Urian - Mihai-Adrian Tecliceanu - Mihai Maganu - Alexandru Ghita
Exposing the Blind Spots: CrowdStrike Research on Feedback-Guided Fuzzing for Comprehensive LLM Testing
June 11, 2025
Paul-Danut Urian - Mihai-Adrian Tecliceanu - Mihai Maganu - Alexandru Ghita AI & Machine LearningCrowdStrike researchers have created a proof-of-concept framework that uses dynamic feedback-guided fuzzing to identify large language model (LLM) vulnerabilities Traditional template-based testing st[…]