Vulnerability report

Exploitation Attempts Observed Confirmed Confidence Not in CISA KEV

CVE-2026-116692

mlflow Remote Code Execution

Mlflow · >= 3.12.0rc0, < 3.17.0

Watch CVE-2026-116692 Browse KEV feed
Severity
CVSS 9.8 · Critical
Confidence
Confirmed
Exploit status
Exploitation Attempts Observed in the Wild
EPSS
0.7%
First observed
28 Oct 2026
Last observed
12 Nov 2026

Decision summary

What security teams need to know first

Direct answers before the deeper technical record.

What it is

CVE-2026-116692 is an unauthenticated vulnerability affecting mlflow. MLflow is an open source AI engineering platform for agents, large language models, and machine learning models.

Have exploitation attempts been observed?

Yes. Previdian sensors have observed 505 exploitation attempts, 28 unique attacker IPs, and 13 countries targeting CVE-2026-116692. First observed Oct 28, 2026; last observed Nov 12, 2026. Confidence is confirmed.

Who is affected?

mlflow >= 3.12.0rc0, < 3.17.0.

What should we do?

Patch immediately, validate internet-facing exposure, and monitor for matching requests.

Overview

MLflow is an open source AI engineering platform for agents, large language models, and machine learning models

MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 3.12.0rc0 until 3.17.0, third-party scorer deserialization in mlflow/genai/scorers/base.py accepts request-selected descendants of THIRD_PARTY_SCORER_ALLOWED_MODULES rather than requiring an exact allowed module. A caller can use MLflow artifact endpoints to place a Python module beneath a writable allowed package and submit its module and class through /ajax-api/3.0/mlflow/scorer/invoke. Scorer.model_validate_json() imports the module and constructs the class before trace processing.

Authentication-disabled deployments permit this code execution without credentials, while basic-auth deployments require a principal capable of creating an experiment and run and uploading artifacts. In either configuration, the allowed scorer package must be writable by the server process. This issue is fixed in version 3.17.0.

mlflow Affected
mlflow
>= 3.12.0rc0, < 3.17.0
Published
27 Oct 2026
Exploitation Attempts Reported
28 Oct 2026
Attack vector
Remote
Complexity
Low
Privileges
None
User interaction
None

Tags

python

CVE References

Exploitation evidence

Evidence of exploitation attempts

Third-party attestation and first-party sensor observation are shown separately so teams can judge the evidence chain.

Exploitation Attempts Observed in the Wild

Previdian

Recorded 28 Oct 2026

Previdian sensors observed exploitation attempts targeting this vulnerability.

Exploitation Attempts Observed

Previdian sensor

First observed 28 Oct 2026

Previdian sensors observed exploitation attempts targeting this vulnerability.

Proof of concept available

GitHub Advisory Database

Recorded 27 Oct 2026

Public scanner or PoC coverage increases practical exploitability.

Known exploited vulnerability sources

Per-source evidence links for KEV attestations are available through the Previdian Pro API.

Learn about Pro API access
Source Added
Previdian Sensors First 2026-11-11 11:16 UTC

Operational indicators for this CVE are listed under Detection.

Sensor telemetry

Previdian Sensor Observations

Previdian sensor observations are exploitation attempts targeting this vulnerability. They do not show that the target was vulnerable or that the attempt succeeded. The cards and chart show volume, unique attackers, and daily activity.

505

Attempts observed

28

Unique attacker IPs

13

Attacker countries

AU · CA · CN · DE · ES · FR · GB · HK · IN · NL · NO · SG · US

4

Sensors observed

CVE-2026-116692 exploitation attempts over the last 17 days

Daily events observed by Previdian sensors

Updated 12 Nov 2026

0
57
111
6
0
7
0
0
0
0
0
0
0
124
0
98
102
27 Oct 31 Oct 4 Nov 8 Nov 12 Nov

First observed 28 Oct 2026 · Last observed 12 Nov 2026

Pro adds sensor region and window summaries. Enterprise adds raw IPs, paths, User-Agents, and payloads.

Start 14-day Pro trial

Detection

Operational artifacts and observed signals

Make the evidence actionable in scanner, SOC, and edge-control workflows.

Observed signals

Request targets
User-Agents
Callback hosts
171
16
2

Request targets and User-Agents available in Pro. Callback host details available in Enterprise.

Scanner coverage

No scanner integrations recorded yet.

Virtual patch status

Previdian virtual patch guidance is available for this CVE.

AWS WAF Cloudflare ModSecurity

Attacker IP Indicators

Attacker IP indicators observed · available in Pro and Enterprise.

Sensor-derived attacker IP indicators are available to Pro and Enterprise accounts under Detection and through the Pro API.

Learn about Pro API access

Virtual Patch Detail

Compensating WAF rules for this CVE.

Available

Enterprise feature. Virtual patch rule content and deployable vendor exports (ModSecurity, Cloudflare, AWS WAF) are available to Previdian Enterprise users.

Talk to Sales Learn About Virtual Patches

Risk and context

Severity, weaknesses, and research context

CVSS v3.1

9.8 Critical

Potential damage if exploited. Separate from whether attackers are using it.

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

EPSS

0.7%

Estimated chance of exploitation in the next 30 days. Previdian's warning comes from evidence, not this score.

Weaknesses

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') CWE-470 Improper Control of Generation of Code ('Code Injection') CWE-94

Recent mention · Github Advisory Database (PIP)

[mlflow] MLflow vulnerable to remote code execution on unauthenticated servers via third-party scorer deserialization importing caller-placed artifact modules

Summary MLflow 3.12.0 introduced registration and deserialization support for third-party scorers in OSS MLflow. During scorer deserialization, the server accepts a request-controlled Python module path beneath one of four allowed package prefixes, imports that module, obtains...

Read full advisory

Potential Proof of Concepts

These PoCs are unverified and could contain malware. Use at your own risk.

https://github.com/mlflow/mlflow/security/advisories/GHSA-26p8-2jq9-3vq9

ghsa · Created Unknown · AI assessment 99%

MLflow vulnerable to remote code execution on unauthenticated servers via third-party scorer deserialization importing caller-placed artifact modules

Timeline

From disclosure to observed exploitation attempts

  1. 01:24 UTC

    Virtual patch available

    Compensating WAF rule available to block exploitation

  2. 05:23 UTC

    Indicators of compromise added (28)

    Indicators of compromise recorded

  3. 05:23 UTC

    Observed by Previdian sensors

    Previdian sensors observed exploitation attempts targeting this vulnerability.

  4. 22:20 UTC

    Added to the Previdian watchlist

    Pro and Enterprise Watch users had 2 weeks of early warning before Previdian KEV.

  5. 22:11 UTC

    Public PoC available

    Public proof-of-concept code published

  6. 20:03 UTC

    CVE published

    Vulnerability disclosed publicly

  7. 20:00 UTC

    CVE ID reserved

    Identifier reserved by the CNA

Pro API

Automate this intelligence

Confidence, exploit status, sensor telemetry, PoCs, scanner integrations, mentions, and tags are available programmatically for VM, SOC, and CTI workflows.

Start 14-day Pro trial Read API docs
  • Evidence confidence and provenance
  • First-party sensor telemetry
  • PoC and scanner references
  • Affected versions and enrichment
  • Automation-ready JSON delivery

GET /api/v2/pro/kevs/CVE-2026-116692

Sample API response 
{
  "cve_id": "CVE-2026-116692",
  "confidence": "Confirmed",
  "cvss_score": 9.8,
  "cvss_estimated": false,
  "epss_score": 0.00704,
  "exploit_status": {
    "exploited_in_the_wild": true,
    "active_exploitation_observed": true
  },
  "sensor_telemetry": { "attempts": 505, "sensors": 4 }
}