arXiv:2608. 05199v1 Announce Type: cross Abstract: Autonomous security agents operate as staged pipelines, such as classifying network traffic and then attributing attacks to a specific technique.
arXiv:2607. 13801v1 Announce Type: cross Abstract: Large language model (LLM)-based intrusion detection systems (IDS) are increasingly studied for security monitoring, yet their robustness against feasible traffic manipulation remains largely empirical.
arXiv:2606. 20502v1 Announce Type: cross Abstract: Whether LLMs scoring well on vulnerability benchmarks genuinely reason about security or merely pattern-match on contaminated data remains unresolved.
arXiv:2606. 29484v1 Announce Type: cross Abstract: Modern deepfake detectors are rarely consumed as bare classifiers.
The paper introduces a provenance‑aware execution graph for long‑horizon LLM agents, defining influence distance (DI) as the shortest structural path from an untrusted source to a sensitive action. Compared to the traditional sequence distance (DT), DI is always less than or equal to DT, revealing a median gap of nine hops in 454 injection–sink pairs across multiple models and datasets. The study shows that most pairs exhibit a non‑zero gap, and a deterministic DI‑based gate can block attacks missed by a sequence‑only gate without extra benign blocking.
arXiv:2607. 18826v1 Announce Type: cross Abstract: LLM-agent defenses are typically evaluated one session at a time.
arXiv:2608. 01388v1 Announce Type: cross Abstract: Runtime safety monitors based on Linear Temporal Logic (LTL) and finite automata (FSA) are increasingly deployed to intercept unsafe tool-call sequences in LLM agents.
DriftNet is a dual‑head trajectory Transformer designed to detect and localize prompt injection attacks in large language model agents. It processes logged tool‑call trajectories, classifying each as compromised or not while labeling every step as benign, injection point, hijacked, or failed injection. On the AgentDrift benchmark, DriftNet achieves high accuracy, with an F1 score of 0.983, 98.7% exact injection‑point recovery, and low false‑alarm rates.
arXiv:2606. 03453v1 Announce Type: cross Abstract: Vulnerability disclosure volumes now far exceed organizational assessment capacity, yet three adjacent research communities (proof-of-concept generation, vulnerability prioritization, and detection rule engineering) operate largely in isolation.
arXiv:2608. 09732v1 Announce Type: cross Abstract: Agent skills are emerging as an important attack surface in LLM-based agent systems.
The paper investigates how to properly validate candidate models before promoting them to replace incumbent classifiers in adaptive network intrusion detection systems. It demonstrates that promotion decisions can be biased by how challengers are constructed and the amount of evidence they receive, and that using self‑contained challenger pipelines and sufficient candidate evidence reduces apparent promotion harm. The study also shows that policy rankings shift with candidate comparability and that no single update policy dominates across benchmarks.
arXiv:2602. 06911v2 Announce Type: replace-cross Abstract: As increasingly capable open-weight large language models (LLMs) are deployed, improving their tamper resistance against unsafe modifications, whether accidental or intentional, becomes critical to minimize risks.
arXiv:2609.06972v1 Announce Type: cross Abstract: LLM agents complete tasks by issuing sequences of tool calls, and every observation they read is a channel through which an indirect prompt injection...