CVE-2026-64849: mlflow · Server-Side Request Forgery (SSRF)
Affects mlflowSource GCVE Database
Exploited in the wild · KEV due 2 Sep 2026 · public detection template (Nuclei) · 10% EPSS : upgrade to 3.15.0 now.
Data as of 26 Oct 2026, 17:16 UTC
- KEV deadline
- US federal agencies must patch by this date; a sensible deadline for everyone else. Other KEV due dates
- EPSS
- 10.0% chance of exploitation in 30 days – higher than 95% of CVEsExploit Prediction Scoring System: the likelihood attackers will use this flaw soon.
- Patch
- Available – fixed in 3.15.0
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:NExploitation evidence
t = 0 is the CVE’s publication on cve.org
Disclosed: evidence found.
CVE published 17 Aug 2026
Detectable: evidence found.
nuclei template: MLflow Webhook SSRF - Unauthenticated Full-Read via Redirect Bypass
Weaponised: no evidence on record.
No public exploit code indexed in Metasploit or Exploit-DB
Exploited in the wildExploited: evidence found.
Listed in CISA KEV 2 days after disclosure
Observed by OffSeq MirageSeen by Mirage: no evidence on record.
No attack traffic recorded by OffSeq Mirage sensors
Timeline
CVE ID reserved
by GitHub, Inc.
, 28 days before disclosure
CVE published
on cve.org
, t = 0
Added to CISA KEV (opens in a new tab)
exploitation confirmed
, 2 days after disclosure
KEV remediation due
US federal deadline
, 16 days after disclosure
A missing step is not evidence of safety. Sources: cve.org (opens in a new tab), CISA KEV (opens in a new tab), the Metasploit, Exploit-DB and nuclei-templates index as of 26 Oct 2026.
What to do
Upgrade mlflow to 3.15.0.
Affected software
| Product | Affected | Fixed in |
|---|---|---|
| mlflow mlflow | Affected
| Fixed in
|
Description
MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. Starting in 3.3.0 and prior to 3.15.0, the unauthenticated POST /api/2.0/mlflow/webhooks/{id}/test endpoint calls _validate_webhook_url() in mlflow/utils/validation.py only for the original URL while mlflow/webhooks/delivery.py follows redirects and re-resolves the hostname without pinning the validated address, allowing attackers to reach internal or cloud metadata services and receive response_status and response_body. This issue is fixed in version 3.15.0.
References
- github.com/mlflow/mlflow/security/advisories/GHSA-7gwp-5pfp-969j (opens in a new tab)
- github.com/mlflow/mlflow/issues/24179 (opens in a new tab)
- github.com/mlflow/mlflow/pull/24258 (opens in a new tab)
- github.com/mlflow/mlflow/commit/ba949522477cbd5915aa55d29b0cfad7d5ddf939 (opens in a new tab)
- github.com/mlflow/mlflow/releases/tag/v3.15.0 (opens in a new tab)
- cisa.gov/known-exploited-vulnerabilities-catalog?field_cve=CVE-2026-64849 (opens in a new tab)