CuraSec

tag: Cache-Poisoning · 1 items

Learn KV Cache Reuse in LLM Reasoning Traces Enables Stealthy Poisoning
2026-10-19 · arXiv cs.CR · source ↗ #llm-security#cache-poisoning#ai-security
  • Engineer — Learn: Novel black-box attack class against LLM inference systems using shared KV caches for reasoning trace reuse — engineers building or deploying such systems should audit whether their architecture exposes shared cache state across users or sessions, and treat this as a design constraint before enabling reasoning-cache reuse features.
  • SOC/IR — Skip
  • Leader — Skip