Research
InjecMEM: Memory Injection Attack on LLM Agent Memory Systems
Publication date not yet evaluated · Added arxiv.org
Page published
Coverage timeline
Single-source research — one report is available.
Why it matters
InjecMEM shows that agent memory subsystems create a persistent attack surface where a single poisoned interaction can silently shape an AI agent's future outputs long after the exchange, unlike transient prompt injection.
InjecMEM, described in a paper by researchers from Shanghai Jiao Tong University and Ant Group, is a memory-injection attack that plants hidden instructions in an LLM agent's persistent memory with a single interaction and no read/edit access to the memory store. Using a retriever-agnostic anchor plus a gradient-optimized adversarial command, it steers later responses on a target topic, achieving up to 35.4% retrieval success and 76.6% attack success on MemoryOS while leaving non-target queries unaffected; code is published on GitHub.