CVE-2026-27497

HIGH
Published February 25, 2026
CISO Take

CVE-2026-27497 is a high-severity RCE in n8n, one of the most widely used AI workflow orchestration platforms. Any authenticated user with workflow editing rights can execute arbitrary code on the n8n server, gaining full control of your AI automation infrastructure and all stored credentials. Patch to 2.10.1/2.9.3/1.123.22 immediately — and treat every n8n instance as potentially compromised until you do.

Affected Systems

Package Ecosystem Vulnerable Range Patched
n8n npm No patch
n8n npm No patch
n8n npm No patch
n8n npm No patch
n8n npm No patch
n8n npm No patch
n8n npm No patch

Severity & Risk

CVSS 3.1
8.8 / 10
EPSS
N/A
KEV Status
Not in KEV
Sophistication
Advanced

Recommended Action

  1. 1. PATCH IMMEDIATELY: Upgrade to n8n 2.10.1, 2.9.3, or 1.123.22. 2. TEMPORARY WORKAROUND (if patching is delayed): Add 'n8n-nodes-base.merge' to the NODES_EXCLUDE environment variable to disable the vulnerable node. 3. RESTRICT PERMISSIONS: Treat workflow creation/editing as an administrative privilege — lock it down to fully trusted users only. 4. AUDIT EXISTING WORKFLOWS: Review all workflows for Merge nodes with suspicious SQL payloads before and after patching. 5. ROTATE CREDENTIALS: After patching, rotate all API keys and service credentials stored in n8n — LLM provider keys, database credentials, cloud tokens. 6. DETECTION: Monitor for unexpected process spawns, file writes outside the n8n data directory, and anomalous outbound connections from the n8n process. Alert on Merge node executions containing SQL-like payloads.

Classification

Compliance Impact

This CVE is relevant to:

EU AI Act
Art.15 - Accuracy, robustness and cybersecurity Article 15 - Accuracy, robustness and cybersecurity Article 9 - Risk management system
ISO 42001
A.6.2.5 - AI system security and safety risks A.8.4 - AI system security A.9.3 - Control of AI system inputs A.9.4 - Access control for AI systems
NIST AI RMF
GOVERN 1.2 - Accountability for AI risk management GOVERN-1.4 - Organizational policies for AI risk MANAGE 2.2 - Mechanisms to respond to identified risks MANAGE-2.2 - Mechanisms to sustain the value of deployed AI
OWASP LLM Top 10
LLM07:2023 - Insecure Plugin Design LLM07:2025 - System Prompt Leakage LLM08:2023 - Excessive Agency LLM08:2025 - Excessive Agency

Technical Details

NVD Description

n8n is an open source workflow automation platform. Prior to versions 2.10.1, 2.9.3, and 1.123.22, an authenticated user with permission to create or modify workflows could leverage the Merge node's SQL query mode to execute arbitrary code and write arbitrary files on the n8n server. The issues have been fixed in n8n versions 2.10.1, 2.9.3, and 1.123.22. Users should upgrade to one of these versions or later to remediate all known vulnerabilities. If upgrading is not immediately possible, administrators should consider the following temporary mitigations. Limit workflow creation and editing permissions to fully trusted users only, and/or disable the Merge node by adding `n8n-nodes-base.merge` to the `NODES_EXCLUDE` environment variable. These workarounds do not fully remediate the risk and should only be used as short-term mitigation measures.

Exploitation Scenario

An insider or compromised developer account with workflow editing rights in n8n creates a new workflow containing a Merge node in SQL query mode. The attacker crafts a payload exploiting CWE-89/CWE-94 to achieve server-side code execution — writing a web shell or spawning a reverse shell. Once on the server, they harvest all n8n-stored credentials: LLM API keys (OpenAI, Anthropic), vector database connections, cloud provider tokens, internal database credentials. These are used to pivot: querying LLMs with exfiltrated proprietary data, poisoning RAG databases, or deploying rogue AI agent workflows that persist under legitimate service accounts. The attack is inherently stealthy because workflow execution logs it as normal n8n activity.

CVSS Vector

CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Timeline

Published
February 25, 2026
Last Modified
March 4, 2026
First Seen
February 25, 2026