CVE-2026-64849 KEV
9.3
CRITICAL · CVSS 3.1 · EPSS 9.8% (pctl 95)
Patch first
On CISA KEV — known exploited in the wild, due 2026-09-02.
Description
MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. 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.
Scoring
| CVSS | 9.3 (CRITICAL, v3.1) |
|---|---|
| Vector | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N |
| EPSS | 9.84% — more likely to be exploited than 95% of all CVEs |
| Weakness | CWE-918 |
| On CISA KEV | yes — remediate by 2026-09-02 |
| Public exploit | none known |
| Published | 2026-08-17 |
| Last modified | 2026-08-20 |
CISA KEV
| Name | MLflow Server-Side Request Forgery Vulnerability |
|---|---|
| Added | 2026-08-19 |
| Due | 2026-09-02 |
| Vendor / product | MLflow / MLflow |
| Ransomware use | none reported |
Affected (1)
| Vendor | Product |
|---|---|
| lfprojects | mlflow |
References
- https://github.com/mlflow/mlflow/commit/ba949522477cbd5915aa55d29b0cfad7d5ddf939
- https://github.com/mlflow/mlflow/issues/24179
- https://github.com/mlflow/mlflow/pull/24258
- https://github.com/mlflow/mlflow/releases/tag/v3.15.0
- https://github.com/mlflow/mlflow/security/advisories/GHSA-7gwp-5pfp-969j
- https://www.cisa.gov/known-exploited-vulnerabilities-catalog?field_cve=CVE-2026-64849
→ the Explorer · watch your stack · NVD