arXiv:2606. 10742v1 Announce Type: cross Abstract: External memory has become a core component of modern web agents, enabling long-horizon reasoning through the retrieval of past experiences.
arXiv:2607. 15657v1 Announce Type: cross Abstract: Multimodal AI agents increasingly rely on persistent long-term memory to ground generation in past visual and textual episodes.
Vis-Poison is a novel attack that poisons multimodal retrieval-augmented generation systems by inserting attacker-controlled images as visual evidence, without altering any textual metadata. The attack uses an automated multi-agent approach to create visually plausible poisoned images and has been tested on two multimodal RAG pipelines, four embedding models, and six generation models. In black-box settings, Vis-Poison achieves an end-to-end success rate between 40.16% and 65.40% against 30,000-entry knowledge bases, and remains effective against various multimodal large language models with an average success rate above 60%.
InjecMEM introduces a memory injection attack that can steer the responses of large language model agents toward a desired output using only a single interaction, without needing read or edit access to the memory store. The attack leverages the retrieval‑then‑generate workflow of memory systems by crafting a retriever‑agnostic anchor with high‑recall topical cues and an adversarial command optimized through gradient‑based coordinate search. Experiments across various memory systems and backbone models show that InjecMEM reliably induces topic‑conditioned retrieval and targeted generation, remains effective even when memory drifts, and does not affect non‑target queries.
Memory is becoming a default subsystem in deployed LLM agents to provide persistent personalization and continuity. This naturally prompts a question: will memory system introduce new vulnerabilities...
arXiv:2607. 05189v1 Announce Type: cross Abstract: Persistent personal agents combine long-term memory with access to users' external environments, enabling personalized foreground assistance and proactive background execution.
arXiv:2606. 15609v1 Announce Type: cross Abstract: Large language model (LLM) agents increasingly rely on long-term memory to support complex task execution, user personalization, and domain adaptation.
arXiv:2606. 04329v1 Announce Type: cross Abstract: Memory is a core component of AI agents, enabling them to accumulate knowledge across interactions and improve performance.
The paper introduces MemJack, a memory‑augmented multi‑agent framework that automatically generates jailbreak attacks on Vision‑Language Models (VLMs) using benign natural images as visual anchors. MemJack discovers visual anchors, camouflages them semantically, evaluates responses, repairs via reflection, and replans dynamically, forming a closed‑loop attack pipeline. The authors also create MemJack‑Bench, a dataset of over 113,000 interactive multimodal jailbreak trajectories, and show that MemJack achieves a 71.48% attack success rate against Qwen3‑VL‑Plus, reaching 90% under extended budgets, outperforming other baselines on natural‑image evaluation.
arXiv:2606. 02643v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG)-enhanced LLM systems, while powerful, introduce substantial inference costs due to the inclusion of an extra multi-stage pipeline that dynamically retrieves and synthesizes information from external knowledge sources.
WAInjectBench introduces the first comprehensive benchmark for detecting prompt injection attacks against web agents, offering a fine‑grained categorization of threats and datasets that include malicious and benign text and image samples. The study systematically evaluates both text‑based and image‑based detection methods across multiple scenarios, revealing that detectors perform well on attacks with explicit instructions or visible perturbations but struggle with subtle or instruction‑free attacks. The authors release the datasets and code to facilitate further research in this area.
Repeat-After-Me is a black-box adaptive visual prompt injection technique that can reveal personally identifiable information or trigger malicious tool calls in both open-weight and commercial vision‑language models, achieving attack success rates above 80% on Qwen3.6‑27B and 47% on GPT‑5.5. The method works even when the benign user prompt is unrelated to the injected task and does not explicitly authorize it, and it retains significant effectiveness when transferred across models or optimized on surrogate systems. In a real‑world OpenClaw Discord deployment, a minimally injected image can overwrite TOOLS.md, enabling remote code execution and secret exfiltration.
arXiv:2607. 06595v1 Announce Type: cross Abstract: Personal AI agents powered by large language models can reason and act using available tools to access emails, manage calendars, and push code to remote repositories, all with minimal oversight.