RadKey: An LLM-Guided RF Backscatter System for Through-Wall Keystroke Inference

In today's digitally connected world, keyboards remain the primary interface for inputting sensitive information, making them a persistent target for eavesdropping attacks. While prior keystroke inference techniques have exploited side-channel signals such as acoustics and vibrations, they...

CVE-2026-31236

A flaw was found in the llm CLI tool. An attacker can exploit a code injection vulnerability by crafting a malicious command with arbitrary Python code in the --functions argument. If a victim is tricked into running this command, it leads to arbitrary code execution on their system, potentially...

CVE-2026-41487

Langfuse is an open source large language model engineering platform. From version 3.68.0 to before version 3.167.0, there is a role-based-access control flaw in the LLM connection update flow. An authenticated, low-privileged user of role “member” in a project could request the update of an...

AI Worm

Researchers have prototyped an AI-powered internet worm. The coolest thing about the prototype is that it carries its own LLM with it, and runs it on computers that have been broken into. This is the closest to John Brunner's original 1975 conception of a computer worm that I've seen...

Beyond Pass/Fail: Using Process Mining to Understand How LLMs Resist (And Fail) Red Team Attacks

Standard AI red teaming evaluations reduce adversarial campaigns to a single binary outcome, attack success rate ASR, not taking into account the sequential structure of how models resist or yield to attacks. We propose applying process mining, a discipline for discovering and analyzing process...

POISE: Position-Aware Undetectable Skill Injection on LLM Agents

Agent skills provide a lightweight mechanism for extending general-purpose agents, but their open format exposes them to skill-poisoning attacks. A practically dangerous injection must stay invisible: if executing the payload derails the user's legitimate task, the resulting failure signal invite...

RedEdit: Agentic Red-Teaming of Image Safety Classifiers Via MCTS-Guided Photo-Editing

Image safety classifiers serve as a critical component of contemporary content moderation systems on the internet. However, their resilience against user-style malicious image editing remains underexplored. Such behaviors are highly prevalent in daily scenarios but difficult to fully reproduce. T...

From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents

Memory is a core component of AI agents, enabling them to accumulate knowledge across interactions and improve performance. However, persistent memory introduces the risk of memory poisoning, where a single adversarial memory write can exert long-term influence over agent behavior. We present a...

-cascade-scan

cascade-scan AI Agent security evaluation framework — autom...

AgentRedBench: Dynamic Redteaming and Integration-Aware Defense for LLM Agents over SaaS Integrations

Indirect prompt injection in tool-use agents is a concrete production threat: LLM agents read from integrations third-party services such as Gmail, Salesforce, or Jira accessed through tool calls whose response content the user neither writes nor controls. Existing benchmarks under-measure the...

Needles at Scale: LLM-Assisted Target Selection for Windows Vulnerability Research

The attack surface of a modern operating system is a haystack: thousands of signed binaries and millions of functions, almost none relevant to any given vulnerability. A human analyst or an LLM agent must pick the function worth reading before analyzing it. At whole-OS scope, this target selectio...

NeuroLog: Reasoning You Can Audit -- Neuro-Symbolic Vulnerability Discovery Via LLM Facts, Datalog, and SMT

Vulnerability discovery on C/C++ source asks the analyst to choose between heavyweight static analysers, which need a working build before a single query runs, and free-form LLMs, which read source readily but invent details and lose track of cross-function dataflow on real codebases. We present...

PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution

Summary The first-party PraisonAI A2A server example combines three behaviors into a remotely exploitable Critical chain: 1. The example exposes an A2A server without configuring authtoken. 2. The same example binds the server to 0.0.0.0. 3. The example registers a calculateexpression tool...

Attackers Use LLM Agent for Post-Exploitation After Marimo CVE-2026-39987 Exploit

An unknown threat actor has been observed using a large language model LLM agent to conduct post-compromise actions after obtaining initial access following the exploitation of a publicly-accessible Marimo network using a recently disclosed vulnerability. "The attacker compromised an...

How to Compare the Security of Code Written by Humans to LLM-Generated Code

Large language models LLMs are rapidly transforming how software is created and maintained. Comparing LLM-generated code against human-written standards is essential to determine whether these new tools uphold or erode the security baselines established by professional developers. Yet, we lack a...

A Protocol-Language Model for Network Intrusion (Without Deep Packet Inspection)

Modern network intrusion detection systems NIDS are caught in a structural contradiction: the protocols carrying the highest threat intelligence are precisely those encrypted under TLS 1.3 and QUIC, where payload inspection yields nothing. We ask a simpler question -- what if the attack signature...

Honeyval: A Comprehensive Evaluation Framework for LLM-Powered HTTP Honeypots

Honeypots are decoy systems mimicking real system components designed to defend against cyber attacks. Recently, LLMs increasingly serve as simulation backbones for honeypots. They enable defenders to construct high-interaction honeypots with low system security risks. However, LLM-powered honeyp...

Automatically Attacking Software Reverse Engineering AI Agents

Software tools for reverse engineering executable binary files, such as Ghidra, enable malware analysts to safely conduct robust static analysis without having access to original source code. Coupled with the analytic power of large language models LLM, agentic systems enabled with tools, such as...

vLLM 安全漏洞

vLLM is an open-source LLM-based inference and service engine that features high throughput and efficient memory usage. Version vLLM 0.14.1 contains a security vulnerability caused by the hardcoding of the trustremotecode=True parameter, which may lead to remote code execution...

Towards Demystifying and Repairing LLM-In-The-Loop Vulnerabilities

Large Language ModelsLLMs have been actively integrated into modern software systems as critical components. LLM-in-the-loop vulnerabilities, where vulnerabilities are introduced by LLMs and their dependent downstream components, such as frameworks, introduce new risks. Although some benchmark...