criticalAgent ThreatPrivilege Abuse

Langflow Unauthenticated Account Creation Leading to RCE

First seen Jul 18, 2026 · Updated Jul 18, 2026 · CVSS 9.8

langflowunauthenticated-accessaccount-creationrcemisconfigurationagent-frameworkASI08 · Cascading FailuresSurface: Tool LayerPropagation: Single Hop

IBM Langflow OSS versions 1.0.0-1.10.0 allow unauthenticated attackers to register new user accounts that, under the common NEW_USER_IS_ACTIVE=true configuration, are immediately active without admin approval. This lets attackers authenticate and reach code-execution endpoints, effectively bypassing intended access controls even when AUTO_LOGIN is disabled. Given the critical CVSS score of 9.8, this is a genuine and severe vulnerability requiring urgent patching.

Technical Analysis

The vulnerability stems from Langflow's user registration endpoint failing to require authentication or authorization before creating new accounts. When the deployment flag NEW_USER_IS_ACTIVE is set to true (a documented and apparently common configuration), these attacker-created accounts are activated instantly rather than requiring admin approval, removing the intended gating mechanism. Once authenticated, the attacker can reach privileged endpoints within Langflow that support flow execution and code/component evaluation, which map to remote code execution (RCE) primitives in the agent-building framework. This crosses the boundary from an unauthenticated network actor to an authenticated internal user with access to agent orchestration and execution logic, effectively granting full compromise of the Langflow instance and any connected agent pipelines, credentials, or downstream tool integrations. The flaw is a classic broken-authentication/authorization issue rather than a novel agentic-AI attack pattern, but its impact is amplified because Langflow is used to construct and run LLM agent workflows with tool access.

Affected Systems

IBM Langflow

Detection Signatures

  • Unexpected POST requests to /api/*/users or registration endpoints from unauthenticated/external IPs
  • Spike in new user account creations without corresponding admin invite or approval logs
  • Newly created accounts authenticating and immediately invoking flow-execution or code-evaluation endpoints
  • Configuration audit: NEW_USER_IS_ACTIVE=true present in environment/deployment manifests
  • Authentication logs showing login shortly after account creation with no email verification step

Remediation Steps

  1. 1

    Upgrade Langflow

    Update to a patched Langflow release beyond 1.10.0 that addresses this CVE as soon as it is available from IBM/Langflow maintainers.

  2. 2

    Disable NEW_USER_IS_ACTIVE

    Set NEW_USER_IS_ACTIVE=false so new registrations require explicit admin activation before the account can authenticate.

  3. 3

    Require authentication for registration

    Restrict or disable open self-registration endpoints; place Langflow behind an authenticated reverse proxy or SSO gateway.

  4. 4

    Audit existing accounts

    Review all user accounts created during the vulnerable window for unauthorized or unrecognized registrations and revoke access.

  5. 5

    Network segmentation

    Ensure Langflow instances are not exposed to the public internet; restrict access to trusted internal networks or VPN.

  6. 6

    Monitor for RCE indicators

    Review logs for flow executions or code-evaluation calls tied to newly created or suspicious accounts.

CVE / Advisory IDs

CVE-2026-9202

Industries Most Exposed

TechnologySoftware DevelopmentAI/ML Platform ProvidersEnterprise ITFinancial ServicesHealthcare

Sources

Respond to this threat

Pro subscribers get a full AI-generated incident-response playbook for this threat — detection, containment, eradication, and recovery steps — plus an unlimited AI Threat Advisor for questions about your environment.