Salami Attack: Stealthy Collusive Memory Poisoning against OpenClaw
Long-term memory enables LLM agents to retain useful information across sessions, but also creates an attack surface through which adversaries may poison an agent's persistent memory to steer its behavior. Existing memory poisoning attacks mainly rely on individually malicious records, overlooking a compositional threat: multiple benign-looking memories may jointly induce unsafe behavior. In this paper, we introduce MemCollusion, an automated red-teaming framework for constructing collusive memo
Record details
Published: 3 August 2026
Source: arXiv red teaming query
Category: Research
Topics: Safety & alignment · Agents & autonomy
Retrieved: 4 August 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
Constructing Executable Analytical Knowledge Representations for Meta-Analysis Synthesis Using an Agentic Harness
arXiv · 3 August 2026
MemSIF: From Structured Interactions to Dual-Track Fact Memory for LLM Agents
arXiv · 3 August 2026
Scoring Rules! Statistical and Strategic Alignment for Text Evaluation Metrics
arXiv cs.LG · 2 August 2026
From Profiling to Synthesis: Benchmarking Implicit Behavioral Alignment in Personalized LLM Agents
arXiv · 3 August 2026
Socially Grounded Agentic AI: Coordinating Plural Perspectives through Social Theory
arXiv · 4 August 2026
OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution
arXiv red teaming query · 1 August 2026
How to cite this record
ethics.ai (3 August 2026), “Salami Attack: Stealthy Collusive Memory Poisoning against OpenClaw,” evidence record 16166, https://ethics.ai/record/16166 (originally published by arXiv red teaming query).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.