CVE-2026-64849
Unauthenticated SSRF in MLflow webhook test allows access to internal/cloud metadata.
Is CVE-2026-64849 being exploited?
Yes. CVE-2026-64849 appears in CISA's Known Exploited Vulnerabilities catalog, which means exploitation in the wild has been confirmed and reported. Treat it as active rather than theoretical, and patch it ahead of higher-scoring vulnerabilities that carry no exploitation evidence.
How severe is CVE-2026-64849?
CVE-2026-64849 is rated Critical with a CVSS score of 9.3. Severity describes how bad exploitation would be, not how likely it is: pair it with exploitation evidence before deciding what to patch first.
Is there a patch for CVE-2026-64849?
Yes. A fix has been recorded for CVE-2026-64849. The vendor advisory is the authority on the exact fixed version — apply it from there rather than from a summary.
What does CVE-2026-64849 affect?
CVE-2026-64849 affects MLflow < 3.15.0, mlflow webhooks test endpoint (/api/2.0/mlflow/webhooks/{id}/test), Deployments with access to cloud metadata services. Confirm the exact affected versions against the vendor advisory before deciding you are exposed.
What should I do about CVE-2026-64849?
Upgrade MLflow to 3.15.0 immediately; block webhook test egress and restrict metadata access.
Exploitation status reflects CISA's KEV catalog as we last synced it. Check the catalog directly.
Unauthenticated SSRF via webhook test endpoint lets attackers query internal/cloud metadata and read response bodies, risking credential and instance compromise.
Immediate action required
- affected< 3.15.0
As published in the CVE Program record. A version outside these ranges is not a statement that it is unaffected — vendors sometimes understate a range, and distribution-backported builds carry upstream numbers that do not reflect what was patched into them.