CVE-2026-103247: n8n: duplicate node IDs bypass credential tamper guard

HIGH
Published October 1, 2026
CISO Take

n8n versions before 1.123.80 contain a credential tampering flaw in which duplicate node IDs defeat the guard that stops workflow editors from reusing credentials they should not control. A low-privileged editor on a shared workflow can keep a victim's stored credentials attached to a node and redirect the secrets to an attacker-controlled host, which turns a limited editor role into theft of API keys and tokens for connected systems. Exploitation likelihood is low for now: EPSS is 0.00236 (87th percentile), the CVE is not in CISA KEV, no public exploit or Nuclei template exists, and SSVC says TRACK. Still, the scope-changed CVSS of 8.5 matters because n8n often holds LLM provider keys, databases and SaaS tokens, and the package has 244 other CVEs. Upgrade to 1.123.80 or later, review who has editor access to shared workflows, and rotate credentials used in workflows that multiple people can edit.

Sources:NVDEPSSGitHub Advisoryvulncheck.comATLASOpenSSFCISA SSVC

What is the risk?

Risk is moderate to high for organizations running multi-user n8n. The CVSS of 8.5 reflects a changed scope with high confidentiality, integrity and availability impact. Exploitation needs network access, a low-privilege authenticated editor account and a complex setup (AC:H), which limits the attacker pool to insiders or compromised accounts. Current exploitation signals are weak: EPSS 0.00236, no KEV listing, no public PoC, no scanner template, and SSVC TRACK. Exposure is concentrated in shared workspaces where credentials are shared across teams. The OpenSSF Scorecard is 6.7/10, and the project's history of 244 CVEs suggests continued vulnerability churn.

How does the attack unfold?

Initial Access
The attacker obtains editor access to a shared n8n workflow through a malicious insider role or a compromised account.
AML.T0012
Guard Bypass
The attacker adds a node with a duplicate ID and a mismatched name so the credential tamper guard fails to strip the victim's credential.
AML.T0083
Credential Exfiltration
The node sends the retained victim credential to an attacker-controlled host when the workflow runs.
AML.T0098
Impact
The attacker reuses the stolen API keys and tokens to access connected AI services, databases and SaaS systems.
AML.T0025

What systems are affected?

PackageEcosystemVulnerable RangePatched
n8nnpm—No patch
210.5KOpenSSF 7.116 dependentsPushed 3d ago42% patched~19d to patchFull package profile →

Do you use n8n? You're affected.

How severe is it?

CVSS 3.1
8.5 / 10
EPSS
0.2%
chance of exploitation in 30 days
Higher than 15% of all CVEs
Exploitation Status
No known exploitation
Sophistication
Moderate

What is the attack surface?

AVACPRUISCIA
AVNetwork
ACHigh
PRLow
UINone
SChanged
CHigh
IHigh
AHigh

What should I do?

1 step
  1. 1) Upgrade n8n to 1.123.80 or later. 2) Until patched, restrict editor access on shared workflows to trusted users and apply least privilege to project and credential sharing. 3) Audit workflows for duplicate node IDs, and for nodes whose credential references or target URLs and hosts changed recently. 4) Rotate credentials, especially LLM and cloud API keys, that were used in workflows editable by multiple users. 5) Apply egress filtering or an allowlist so n8n can only reach approved hosts, and alert on outbound requests to unknown domains. 6) Review n8n audit logs for workflow edits by non-owners, and monitor for unexpected use of provider keys.

What does CISA's SSVC say?

DecisionTrack
Exploitationnone
AutomatableNo
Technical Impacttotal

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act
Art. 15-Accuracy, robustness and cybersecurity
ISO 42001
A.6.2.6-AI system operation and monitoring
NIST AI RMF
MANAGE 2.3-Respond to previously unknown risks
OWASP LLM Top 10
LLM06-Excessive Agency

Frequently Asked Questions

What is CVE-2026-103247?

n8n versions before 1.123.80 contain a credential tampering flaw in which duplicate node IDs defeat the guard that stops workflow editors from reusing credentials they should not control. A low-privileged editor on a shared workflow can keep a victim's stored credentials attached to a node and redirect the secrets to an attacker-controlled host, which turns a limited editor role into theft of API keys and tokens for connected systems. Exploitation likelihood is low for now: EPSS is 0.00236 (87th percentile), the CVE is not in CISA KEV, no public exploit or Nuclei template exists, and SSVC says TRACK. Still, the scope-changed CVSS of 8.5 matters because n8n often holds LLM provider keys, databases and SaaS tokens, and the package has 244 other CVEs. Upgrade to 1.123.80 or later, review who has editor access to shared workflows, and rotate credentials used in workflows that multiple people can edit.

Is CVE-2026-103247 actively exploited?

No confirmed active exploitation of CVE-2026-103247 has been reported, but organizations should still patch proactively.

How to fix CVE-2026-103247?

1) Upgrade n8n to 1.123.80 or later. 2) Until patched, restrict editor access on shared workflows to trusted users and apply least privilege to project and credential sharing. 3) Audit workflows for duplicate node IDs, and for nodes whose credential references or target URLs and hosts changed recently. 4) Rotate credentials, especially LLM and cloud API keys, that were used in workflows editable by multiple users. 5) Apply egress filtering or an allowlist so n8n can only reach approved hosts, and alert on outbound requests to unknown domains. 6) Review n8n audit logs for workflow edits by non-owners, and monitor for unexpected use of provider keys.

What systems are affected by CVE-2026-103247?

This vulnerability affects the following AI/ML architecture patterns: Agent frameworks, Workflow automation pipelines, LLM orchestration, RAG pipelines, Multi-tenant shared workspaces.

What is the CVSS score for CVE-2026-103247?

CVE-2026-103247 has a CVSS v3.1 base score of 8.5 (HIGH). The EPSS exploitation probability is 0.24%.

What is the AI security impact?

Affected AI Architectures

Agent frameworksWorkflow automation pipelinesLLM orchestrationRAG pipelinesMulti-tenant shared workspaces

MITRE ATLAS Techniques

AML.T0012Valid Accounts
AML.T0025Exfiltration via Cyber Means
AML.T0083Credentials from AI Agent Configuration
AML.T0098AI Agent Tool Credential Harvesting

Compliance Controls Affected

EU AI Act: Art. 15
ISO 42001: A.6.2.6
NIST AI RMF: MANAGE 2.3
OWASP LLM Top 10: LLM06

What are the technical details?

Original Advisory

n8n versions before 1.123.80 contain a credential tampering vulnerability where duplicate node IDs bypass the workflow credential tamper guard. Attackers with editor access to shared workflows can exploit mismatched node ID and name matching to retain victim credentials and redirect secrets to attacker-controlled hosts.

Exploitation Scenario

An attacker holds editor access to a shared n8n project, either as a malicious insider or through a compromised account. They edit a workflow that uses a colleague's OpenAI and database credentials, and add a node whose ID duplicates an existing node while its name differs. The duplicate ID defeats the tamper guard, so the node keeps the victim's credential binding while its target URL points to an attacker-controlled server. When the workflow runs, n8n sends the authenticated request with the victim's secret to the attacker's host. The attacker then uses the stolen keys to reach LLM services and connected data stores outside n8n.

Weaknesses (CWE)

CWE-639 — Authorization Bypass Through User-Controlled Key: The system's authorization functionality does not prevent one user from gaining access to another user's data or record by modifying the key value identifying the data.

  • [Architecture and Design] For each and every data access, ensure that the user has sufficient privilege to access the record that is being requested.
  • [Architecture and Design, Implementation] Make sure that the key that is used in the lookup of a specific user's record is not controllable externally by the user or that any tampering can be detected.

Source: MITRE CWE corpus.

CVSS Vector

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

Timeline

Published
October 1, 2026
Last Modified
October 1, 2026
First Seen
October 1, 2026

Related Vulnerabilities