CVE-2024-37061: MLflow RCE — HIGH

CISO Take

Any MLflow deployment running version 1.11.0 or later is vulnerable to arbitrary code execution if a user runs a maliciously crafted MLproject file. This is a realistic social engineering vector in ML teams that routinely share experiment configurations via repos or model hubs. Patch immediately and audit who can submit MLproject files to shared ML platforms.

What is the risk?

HIGH risk for organizations with active MLflow deployments. CVSS 8.8 reflects low attack complexity and no privilege requirement — only user interaction needed, which is trivially achieved in ML environments where sharing MLproject configs is standard practice. The attack surface expands significantly in collaborative ML platforms, shared experiment trackers, and CI/CD pipelines that auto-execute MLprojects from repos. No KEV listing yet, but the exploit reference from HiddenLayer indicates public PoC availability.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
MLflow	pip	—	No patch
26.6K OpenSSF 5.6 655 dependents Pushed 4d ago 31% patched ~51d to patch Full package profile →

Do you use MLflow? You're affected.

How severe is it?

CVSS 3.1

8.8 / 10

EPSS

0.9%

chance of exploitation in 30 days

Higher than 54% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Trivial

Exploitation Confidence

medium

○ CISA SSVC: Public PoC

○ Public PoC indexed (trickest/cve)

Composite signal derived from CISA KEV, VulnCheck KEV, CISA SSVC, EPSS, Metasploit, Exploit-DB, trickest/cve, Nuclei templates, and inthewild.io exploitation reports.

What is the attack surface?

AV Network

AC Low

PR None

UI Required

S Unchanged

C High

I High

A High

What should I do?

6 steps

PATCH

Upgrade MLflow to the latest patched version immediately. Verify via pip show mlflow.
RESTRICT

Block execution of MLproject files from untrusted sources at the process and policy level. Implement allowlists for MLproject sources.
SANDBOX

Run mlflow run in isolated containers with no network access and minimal filesystem permissions.
AUDIT

Review CI/CD pipelines that auto-execute MLproject files — add approval gates for any externally sourced projects.
DETECT

Alert on unexpected outbound connections or process spawns from MLflow worker processes. Monitor for mlflow run executions against non-internal URIs.
REVIEW

Audit shared MLproject files in internal repos for malicious code patterns, especially in entry_points and conda.yaml/requirements.txt sections.

What does CISA's SSVC say?

Decision Attend

Exploitation poc

Automatable No

Technical Impact total

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

How is it classified?

Code Execution Supply Chain Framework Training Data AML.T0010.001 - AI Software AML.T0011 - User Execution AML.T0011.000 - Unsafe AI Artifacts AML.T0049 - Exploit Public-Facing Application AML.T0050 - Command and Scripting Interpreter AML.T0072 - Reverse Shell

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Art.15 - Accuracy, robustness and cybersecurity Art.9 - Risk management system

ISO 42001

6.1.2 - AI risk assessment 8.4 - AI system supply chain

NIST AI RMF

GOVERN-1.2 - Policies for AI risk management MANAGE-2.2 - Mechanisms for AI risk response

OWASP LLM Top 10

LLM03:2025 - Supply Chain

Frequently Asked Questions

What is CVE-2024-37061?

Any MLflow deployment running version 1.11.0 or later is vulnerable to arbitrary code execution if a user runs a maliciously crafted MLproject file. This is a realistic social engineering vector in ML teams that routinely share experiment configurations via repos or model hubs. Patch immediately and audit who can submit MLproject files to shared ML platforms.

Is CVE-2024-37061 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2024-37061, increasing the risk of exploitation.

How to fix CVE-2024-37061?

1. PATCH: Upgrade MLflow to the latest patched version immediately. Verify via `pip show mlflow`. 2. RESTRICT: Block execution of MLproject files from untrusted sources at the process and policy level. Implement allowlists for MLproject sources. 3. SANDBOX: Run `mlflow run` in isolated containers with no network access and minimal filesystem permissions. 4. AUDIT: Review CI/CD pipelines that auto-execute MLproject files — add approval gates for any externally sourced projects. 5. DETECT: Alert on unexpected outbound connections or process spawns from MLflow worker processes. Monitor for `mlflow run` executions against non-internal URIs. 6. REVIEW: Audit shared MLproject files in internal repos for malicious code patterns, especially in entry_points and conda.yaml/requirements.txt sections.

What systems are affected by CVE-2024-37061?

This vulnerability affects the following AI/ML architecture patterns: training pipelines, MLOps platforms, experiment tracking, CI/CD ML pipelines, model registries.

What is the CVSS score for CVE-2024-37061?

CVE-2024-37061 has a CVSS v3.1 base score of 8.8 (HIGH). The EPSS exploitation probability is 0.88%.

What is the AI security impact?

Affected AI Architectures

training pipelinesMLOps platformsexperiment trackingCI/CD ML pipelinesmodel registries

MITRE ATLAS Techniques

AML.T0010.001 AI Software

AML.T0011 User Execution

AML.T0011.000 Unsafe AI Artifacts

AML.T0049 Exploit Public-Facing Application

AML.T0050 Command and Scripting Interpreter

AML.T0072 Reverse Shell

Compliance Controls Affected

EU AI Act: Art.15, Art.9

ISO 42001: 6.1.2, 8.4

NIST AI RMF: GOVERN-1.2, MANAGE-2.2

OWASP LLM Top 10: LLM03:2025

What are the technical details?

Original Advisory

Remote Code Execution can occur in versions of the MLflow platform running version 1.11.0 or newer, enabling a maliciously crafted MLproject to execute arbitrary code on an end user’s system when run.

Exploitation Scenario

Attacker publishes a weaponized MLproject to a public GitHub repository or submits it to an internal ML experiment tracker. The MLproject's entry_points section contains a malicious command disguised as a training script invocation. A data scientist or automated CI/CD pipeline runs `mlflow run <malicious-repo>`. MLflow clones the project and executes the entry point, triggering arbitrary code execution on the host. From there, the attacker can dump cloud credentials from the execution environment (common in AWS/GCP ML workloads), exfiltrate model weights and training data, or establish persistence in the MLOps infrastructure. The attack is especially effective against teams that routinely reproduce experiments from external sources.

Weaknesses (CWE)

CWE-94 Improper Control of Generation of Code ('Code Injection') Primary CWE-94 Improper Control of Generation of Code ('Code Injection')

CWE-94 — Improper Control of Generation of Code ('Code Injection'): The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.

[Architecture and Design] Refactor your program so that you do not have to dynamically generate code.
[Architecture and Design] Run your code in a "jail" or similar sandbox environment that enforces strict boundaries between the process and the operating system. This may effectively restrict which code can be executed by your product. Examples include the Unix chroot jail and AppArmor. In general, managed code may provide some protection. This may not be a feasible solution, and it only limits the impact to the operating system; the rest of your application may still be subject to compromise. Be careful to avoid CWE-243 and other weaknesses related to jails.

Source: MITRE CWE corpus.