The Vulnerability with No CVE: Managing Persistent Gaps between Mandate and Authority in AI Coding Agents

Existing guidance identifies excessive agency, excessive permission, weak task-bound authorization, and inadequate agent controls as important risks. Control frameworks also describe capabilities for constraining, authorizing, observing, validating, and responding to agent activity. Yet security...

Incident Report: Unsanctioned Agent Behaviour during Cyber Testing

During a routine cyber evaluation, AISI identified an incident in which AI agents took sustained, unsanctioned action directed at real people and organisations. The AI Security Institute is disclosing what they found, what it means, and the actions now underway...

EUVD-2026-42609

n8n: "Allowed HTTP Request Domains" Restriction Bypass via AI Agents MCP Connector...

GHSA-X5VX-C2C8-M3W9 n8n: AI Agents Project Viewer Privilege Escalation via run_node_tool

Impact In n8n's AI Agents feature, a user with the read-only Project Viewer role could escalate their privileges by chatting with an agent that has node tools enabled. The agent's node-execution tool was authorized only by the agent:execute scope and ran nodes using the project's credentials,...

GHSA-W46P-W7W2-FR9G Duplicate Advisory: AI Agents Project Viewer Privilege Escalation via run_node_tool

Duplicate Advisory This advisory has been withdrawn because it is a duplicate of GHSA-x5vx-c2c8-m3w9. This link is maintained to preserve external references. Original Description n8n versions before 2.30.1 contain a privilege escalation vulnerability in the AI Agents feature where the...

CVE-2026-65015 n8n before 2.30.1 Privilege Escalation via run_node_tool

n8n versions before 2.30.1 contain a privilege escalation vulnerability in the AI Agents feature where the node-execution tool lacks proper authorization checks. A Project Viewer user can escalate privileges by chatting with an agent that has node tools enabled, executing arbitrary nodes and...

PT-2026-63345

Name of the Vulnerable Software and Affected Versions n8n versions prior to 2.30.1 Description A privilege escalation issue exists in the AI Agents feature where the node-execution tool lacks proper authorization checks. A user with the Project Viewer role can escalate privileges by interacting...

Open-Source Android AI Agents Could Let Invisible Screen Text Run Code on Host PCs

An Android app that can draw over other windows and write to shared storage can slip instructions to the AI agent driving that phone, in text no human eye will ever see. Two more steps, and the same app is running commands on the PC driving the agent. Researchers demonstrated that chain, plus six...

From Recon to Free Flights: Precision Prompt Attacks on AI Agents

...

Least privilege for AI agents: Identity, access, and tool binding

In this article 1. Real-world scenarios 2. Best Practices: Identity + RBAC + Scope + Safe Tool Binding 3. Looking Ahead AI agents aren't only smarter API callers. They plan, chain actions across systems, and invoke tools in sequences while no single human explicitly approves each step. The...

Least privilege for AI agents: Identity, access, and tool binding

In this article 1. Real-world scenarios 2. Best Practices: Identity + RBAC + Scope + Safe Tool Binding 3. Looking Ahead AI agents aren't only smarter API callers. They plan, chain actions across systems, and invoke tools in sequences while no single human explicitly approves each step. The...

EUVD-2026-43061

Incorrect Authorization vulnerability in Drupal AI Agents allows Forceful Browsing. This issue affects AI Agents versions: from 0.0.0 to 1.1.4, from 1.2.0 to 1.2.5, from 1.3.0 to 1.3.1...

AI Coding Agents Found Triggering Endpoint Security Rules Built to Catch Attackers

Sophos looked at a week of its own endpoint data and found that AI coding agents such as Claude Code, Cursor, and OpenAI Codex are setting off detection rules written to catch human intruders. The agents are not malicious. They just do a lot of things that, to a behavioral engine, look exactly li...

SkillCloak Lets Malicious AI Agent Skills Evade Static Scanners with Self-Extracting Packing

Scanners meant to catch malicious add-on "skills" for AI coding agents can be fooled by a few simple changes that leave the malware working, according to a new study from researchers at the Hong Kong University of Science and Technology. Their strongest trick slipped past every scanner tested mor...

Identity Lifecycle Management Wasn't Built for AI Agents 

Identity lifecycle management was architected around a person with an employment record, a manager, and a departure date. AI agents have none of those. As autonomous principals proliferate across enterprise environments, the governance model built for humans develops structural blind spots that...

OpenClaw: risks for the users and how to mitigate them

OpenClaw, which was previously known as Clawdbot and Moltbot, is today one of the most successful and fast‑growing ecosystems for AI agents, recognized worldwide. The project quickly became popular with users because of its flexibility and ability to solve fairly complex tasks that previously...

Understanding the (In)Security of Vibe-Coded Applications

Recent advances in large language models LLMs have enabled vibe coding, an emerging software development paradigm in which users create applications primarily through natural-language interactions with AI agents. Due to its low barrier to entry, vibe coding is rapidly gaining adoption in practice...

Forget Data Leakage: Shadow AI's Real Threat Is Access Control

The first wave of enterprise AI concern was straightforward. It was simply employees pasting sensitive data into public AI tools. Security teams responded with usage policies, domain blocks, and data loss prevention rules. That response made sense at the time. It doesn't fit the problem anymore...

MAL-2026-5563 Malicious code in @sentry-internal-sdk/profiling-node (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector c7951165844874f57819b0d63b8c8511e4e9217bf0f9231ec02f06cb6e059c47 Package name @sentry-internal-sdk/profiling-node impersonates the legitimate @sentry/profiling-node Sentry publishes under the @sentry org; no...

Lessons from Penetration Tests on Large-Scale Agent Systems

As AI systems gain increasing autonomy and execution capability, the number of discovered security vulnerabilities continues to rise. However, many of these vulnerabilities are not fundamentally novel, but instead reflect recurring classes of weaknesses long observed in prior computing systems...