Skip to main content
Vulnerability Database/CVE-2026-97680

CVE-2026-97680: IBM Langflow OSS Information Disclosure

CVE-2026-97680 is an information disclosure flaw in IBM Langflow OSS versions 1.0.0 through 1.12.2 that enables authenticated attackers to access sensitive data or inject malicious content. This article covers technical details, affected versions, impact assessment, and mitigation strategies.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-97680 Overview

CVE-2026-97680 is an improper access control vulnerability [CWE-284] in IBM Langflow OSS versions 1.0.0 through 1.12.2. The flaw resides in the vertex result caching subsystem. A remote authenticated attacker can exploit the weakness to read sensitive information or inject malicious data into cached results. Successful exploitation compromises the confidentiality and integrity of cached vertex outputs shared across the Langflow runtime.

Critical Impact

Authenticated attackers can access or tamper with cached vertex results belonging to other users or flows, enabling information disclosure and data manipulation across tenants.

Affected Products

  • IBM Langflow OSS 1.0.0 through 1.12.2
  • Deployments exposing the Langflow runtime to multiple authenticated users
  • Environments using the vertex result caching subsystem

Discovery Timeline

  • 2026-10-07 - CVE-2026-97680 published to NVD
  • 2026-10-07 - Last updated in NVD database

Technical Details for CVE-2026-97680

Vulnerability Analysis

IBM Langflow OSS caches vertex execution results to accelerate reuse across flow runs. The caching subsystem fails to enforce proper access control checks when retrieving or writing cached entries. An authenticated user can therefore reach cached results produced in contexts to which they should not have access. The same gap permits writing attacker-controlled values into cache entries that downstream consumers trust.

The weakness falls under improper access control [CWE-284]. It affects both confidentiality and integrity of workflow execution, and it degrades availability when poisoned cache entries disrupt flow runs. Attack complexity is low and no user interaction is required, which raises the practical exposure for multi-tenant Langflow deployments.

Root Cause

The vertex result caching layer does not consistently validate the requesting principal against the owner or scope of the cached entry. Authorization decisions rely on cache key possession rather than on an identity or tenant check. Any authenticated caller who can reach the caching endpoints and supply or guess a cache identifier can read or overwrite the stored value.

Attack Vector

Exploitation is network-based and requires only low-privilege authenticated access. The attacker interacts with the Langflow API endpoints that mediate vertex result caching. By issuing read requests against cache identifiers owned by other users, the attacker obtains sensitive inputs, outputs, or secrets embedded in flow data. By issuing write requests, the attacker injects malicious payloads that subsequent flow executions consume as trusted results.

No verified public exploit code is available for CVE-2026-97680. Refer to the IBM Support Page for authoritative technical details.

Detection Methods for CVE-2026-97680

Indicators of Compromise

  • Unexpected read or write requests against Langflow vertex cache endpoints originating from low-privilege accounts.
  • Flow execution outputs that diverge from expected results, suggesting cache poisoning of vertex entries.
  • Access to cached vertex identifiers that do not belong to the authenticated user's flows or workspace.

Detection Strategies

  • Enable verbose API access logging on Langflow and alert on cross-user access patterns to vertex cache resources.
  • Correlate authenticated session identity with the owner of each cache key accessed and flag mismatches.
  • Hunt for anomalous bursts of cache read operations by a single authenticated principal across many flow or vertex identifiers.

Monitoring Recommendations

  • Track all HTTP requests to Langflow caching routes with request body, response size, and principal metadata.
  • Baseline normal cache access volume per user and alert on deviations.
  • Review audit logs for authenticated accounts created or modified shortly before suspicious cache activity.

How to Mitigate CVE-2026-97680

Immediate Actions Required

  • Upgrade IBM Langflow OSS to a fixed version as published on the IBM Support Page.
  • Restrict network access to Langflow instances so that only trusted authenticated users can reach the API.
  • Rotate credentials, API keys, and secrets that may have been exposed through cached vertex results.
  • Audit existing cache contents and flow definitions for signs of tampering before restoring normal operations.

Patch Information

IBM has published remediation guidance for CVE-2026-97680 on the vendor advisory. Review the IBM Support Page for the fixed version matrix and upgrade procedure. Apply the upgrade across all Langflow OSS deployments running versions 1.0.0 through 1.12.2.

Workarounds

  • Disable or constrain the vertex result caching subsystem until upgrade is possible, accepting the performance tradeoff.
  • Segment Langflow deployments per tenant or team so that cache contents are not shared across trust boundaries.
  • Enforce least-privilege authentication and remove unused Langflow accounts to reduce the pool of potential attackers.
bash
# Example: verify installed Langflow version before and after patching
pip show langflow | grep -i version

# Example: upgrade Langflow OSS to the fixed release
pip install --upgrade "langflow>FIXED_VERSION"

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.