Cloud agents typically send user messages to remote LLMs and store conversation traces in memory systems (e.g., Mem0, LangMem, Memobase) for long-term personalization. This creates a large privacy attack surface: plaintext prompts and logs may contain PII, medical/financial data, credentials cloud memory stores can leak via retrieval, prompt injection, inversion, or misconfiguration naïve mitigation (e.g., *** masking) destroys task semantics, harming retrieval and personalization Goal: reduce privacy leakage without sacrificing utility.
Paper page - MemPrivacy: Privacy-Preserving Personalized Memory Management for Edge-Cloud AgentsPapers arxiv:2604.28157 FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption Published on Apr 30 Submitted by yanting wang on May 1 Pennsylvania State University Authors: , , , Abstract FlashRT is a framework that significantly improves the… …
… The following papers were recommended by the Semantic Scholar API When Background Matters: Breaking Medical Vision Language Models by Transferable Attack 2026 Towards Understanding the Robustness of Sparse Autoencoders 2026 Adversarial Prompt Injection Attack on Multimodal Large Language Models 202… …
… Defending state-of-the-art multi-turn malicious attacks like CKA-Agent . …
… The following papers were recommended by the Semantic Scholar API PISmith: Reinforcement Learning-based Red Teaming for Prompt Injection Defenses 2026 Uncovering Linguistic Fragility in Vision-Language-Action Models via Diversity-Aware Red Teaming 2026 T-MAP: Red-Teaming LLM Agents with Trajectory-… …
… This creates a large privacy attack surface: plaintext prompts and logs may contain PII , medical/financial data, credentials cloud memory stores can leak via retrieval, prompt injection, inversion, or misconfiguration naïve mitigation e.g., masking destroys task semantics , harming retrieval and p… …