CVE-2026-97671: Langflow: path traversal exposes sensitive files

MEDIUM

Published October 7, 2026

CISO Take

IBM Langflow OSS 1.0.0 through 1.12.2 contains a path traversal flaw (CWE-22) that lets any authenticated user read files outside the intended directory, rated medium (CVSS 6.5) with high confidentiality impact. Exploitation probability is low for now: EPSS is 0.446% (63rd percentile), the CVE is not in CISA KEV, and there is no public exploit or Nuclei template. Still, Langflow has a poor security track record, with 194 other CVEs in the same package and a package risk score of 77/100, and low-privilege credentials are easy to obtain on shared or weakly secured instances. Langflow hosts often hold LLM API keys, database credentials and flow definitions, so file read can become a stepping stone to wider compromise. Upgrade to the fixed release listed in the IBM advisory (https://www.ibm.com/support/pages/node/7290694), since no patched version is published in our data, and restrict the instance to trusted users and networks until you do.

Sources:NVDEPSSATLASibm.com

What is the risk?

Medium overall. The attack is network-reachable, low complexity, needs no user interaction and requires only low privileges, with high confidentiality impact and no integrity or availability impact. Exploitation likelihood is currently low (EPSS 0.00446, no public exploit, not in KEV). Risk rises sharply for internet-exposed instances, multi-tenant deployments, or those with open self-registration, and for hosts that store secrets in environment files or on disk. Langflow's 194 prior CVEs suggest attackers and researchers already focus on this product.

How does the attack unfold?

Initial Access

Attacker authenticates to the Langflow instance with a low-privilege or self-registered account.

AML.T0049

Exploitation

Attacker sends a request containing path traversal sequences to a file-handling endpoint to escape the intended directory.

AML.T0037

Credential Collection

Attacker reads .env files and configs to harvest LLM API keys and database credentials.

AML.T0055

Impact

Stolen secrets and flow data enable abuse of model APIs, access to RAG data stores, and loss of proprietary logic.

Initial Access

Attacker authenticates to the Langflow instance with a low-privilege or self-registered account.

AML.T0049

Exploitation

Attacker sends a request containing path traversal sequences to a file-handling endpoint to escape the intended directory.

AML.T0037

Credential Collection

Attacker reads .env files and configs to harvest LLM API keys and database credentials.

AML.T0055

Impact

Stolen secrets and flow data enable abuse of model APIs, access to RAG data stores, and loss of proprietary logic.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
Langflow	pip	—	No patch
156.0KPushed 6d ago18% patched~66d to patchFull package profile →

Do you use Langflow? You're affected.

How severe is it?

CVSS 3.1

6.5 / 10

EPSS

0.5%

chance of exploitation in 30 days

Higher than 43% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

No known exploitation

Sophistication

Trivial

What is the attack surface?

AVNetwork

ACLow

PRLow

UINone

SUnchanged

CHigh

INone

ANone

What should I do?

1 step

1) Upgrade Langflow to the fixed version named in the IBM advisory (https://www.ibm.com/support/pages/node/7290694); our data lists no patched version, so confirm it with the vendor. 2) Until patched, put Langflow behind a VPN or authenticating reverse proxy, disable open registration, and limit accounts to trusted users. 3) Run the service as a low-privilege user in a container with a read-only filesystem and only the needed volumes mounted, so the process cannot reach host secrets. 4) Move secrets out of local files into a secrets manager and rotate any keys stored on affected hosts. 5) Detect exploitation by watching web and application logs for ../, ..%2f, %2e%2e and absolute-path patterns in file-related API requests, and for unusual reads of sensitive files by the Langflow process.

How is it classified?

Data LeakageData ExtractionFrameworkAPIAML.T0037- Data from Local System AML.T0049- Exploit Public-Facing Application AML.T0055- Unsecured Credentials

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Art. 15-Accuracy, robustness and cybersecurity

ISO 42001

A.10.3-Suppliers

NIST AI RMF

MANAGE 2.3-Responding to unknown risks and managing third-party AI components

OWASP LLM Top 10

LLM02:2025-Sensitive Information DisclosureLLM03:2025-Supply Chain

Frequently Asked Questions

What is CVE-2026-97671?

Is CVE-2026-97671 actively exploited?

No confirmed active exploitation of CVE-2026-97671 has been reported, but organizations should still patch proactively.

How to fix CVE-2026-97671?

1) Upgrade Langflow to the fixed version named in the IBM advisory (https://www.ibm.com/support/pages/node/7290694); our data lists no patched version, so confirm it with the vendor. 2) Until patched, put Langflow behind a VPN or authenticating reverse proxy, disable open registration, and limit accounts to trusted users. 3) Run the service as a low-privilege user in a container with a read-only filesystem and only the needed volumes mounted, so the process cannot reach host secrets. 4) Move secrets out of local files into a secrets manager and rotate any keys stored on affected hosts. 5) Detect exploitation by watching web and application logs for ../, ..%2f, %2e%2e and absolute-path patterns in file-related API requests, and for unusual reads of sensitive files by the Langflow process.

What systems are affected by CVE-2026-97671?

This vulnerability affects the following AI/ML architecture patterns: Agent frameworks, LLM workflow orchestration, RAG pipelines, Low-code AI application builders.

What is the CVSS score for CVE-2026-97671?

CVE-2026-97671 has a CVSS v3.1 base score of 6.5 (MEDIUM). The EPSS exploitation probability is 0.50%.

What is the AI security impact?

Affected AI Architectures

Agent frameworksLLM workflow orchestrationRAG pipelinesLow-code AI application builders

MITRE ATLAS Techniques

AML.T0037Data from Local System

AML.T0049Exploit Public-Facing Application

AML.T0055Unsecured Credentials

Compliance Controls Affected

EU AI Act: Art. 15

ISO 42001: A.10.3

NIST AI RMF: MANAGE 2.3

OWASP LLM Top 10: LLM02:2025, LLM03:2025

What are the technical details?

Original Advisory

IBM Langflow OSS 1.0.0 through 1.12.2 could allow a remote authenticated attacker to obtain sensitive information due to a path traversal vulnerability.

Exploitation Scenario

An attacker obtains a low-privilege Langflow account, for example through self-registration, a leaked credential or a shared team login. They send requests to a file-handling endpoint with traversal sequences such as ../../ to escape the intended directory. They read the .env file, configuration files and database credentials on the host, recovering OpenAI or Anthropic API keys and vector database tokens. With those, they query the organisation's models and data stores directly, run up costs, and extract proprietary documents used in RAG pipelines.

Weaknesses (CWE)

CWE-22Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')Primary CWE-22Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')

CWE-22 — Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal'): The product uses external input to construct a pathname that is intended to identify a file or directory that is located underneath a restricted parent directory, but the product does not properly neutralize special elements within the pathname that can cause the pathname to resolve to a location that is outside of the restricted directory.

[Implementation] Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does. When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, "boat" may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as "red" or "blue." Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code's environment changes. This can give attackers enough room to bypass the intended validation. However, denylis
[Architecture and Design] For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks entirely. Then, these modified values would be submitted to the server.

Source: MITRE CWE corpus.