Continuous Discovery of Vulnerabilities in LLM Serving Systems with Fuzzing

LLM inference and serving systems have become security-critical infrastructure; however, many of their most concerning failures arise from the serving layer rather than from model behavior alone. Modern inference engines combine KV cache, batching, prefix sharing, speculative decoding, adapters,...

CVE-2025-31534

creationtimestamp| type| source ---|---|--- 2025-04-01 21:40:30+00:00| seen| https://bsky.app/profile/cyberalerts.bsky.social/post/3llrtleuvb722 2025-04-01 23:48:39+00:00| seen| https://mastodon.social/users/CyberSignaler/statuses/114265379121005167 2025-04-01 23:48:39+00:00| seen|...