Skip to content
NewA calmer, faster ZeroHour

Authenticated MLflow webhook SSRF rated CVSS 6.8 (CVE-2026-64849)

Original title: ZDI-26-846: MLflow Webhook Server-Side Request Forgery Vulnerability

ZDI Published Advisories ()1 · 1 read
Part of a story covered by 9 sourcesZDI publishes nine MLflow advisories; fixes and exploitation unstatedMerged summary & timeline
AI summary · x-ai/grok-4.7Summaries by Vani AI

Logged-in attackers can make the server send arbitrary requests; versions and a fix are not given.

ZDI published ZDI-26-846 for a server-side request forgery flaw in MLflow webhooks. Remote attackers who are already authenticated can initiate arbitrary server-side requests. ZDI assigned CVSS 6.8 and CVE-2026-64849. The snippet does not list versions, a patch, or exploitation.

Unconfirmed: Affected versions, remediation, and exploitation status are not stated.

Key facts
  • Authentication is required to exploit the webhook flaw.
  • ZDI assigned a CVSS rating of 6.8.
  • The advisory is ZDI-26-846 and the CVE is CVE-2026-64849.
  • No fixed version is stated in the snippet.
Security assessment · AI, from the article

Impact: Authenticated attackers can initiate arbitrary server-side requests

ProductAffected
MLflow—
Analyze CVE-2026-64849 with Vani AIImpact, affected versions, exploitation, mitigations and detection

Vulnerabilities mentioned

All
CVEVulnerabilityCVSSEPSSSignalsPublished
CVE-2026-64849
Unauthenticated SSRF in MLflow Exposes Cloud Credentials and Secrets
9.310%KEVPoC ×5
KEV

From the original article

This vulnerability allows remote attackers to initiate arbitrary server-side requests on affected installations of MLflow. Authentication is required to exploit this vulnerability. The ZDI has assigned a CVSS rating of 6.

ZDI Published Advisories

zerodayinitiative.com

Read the full article at ZDI Published Advisories

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.