MLflow Vulnerability Exploited for Cloud Credential Theft
The critical-severity flaw allows attackers to send HTTP requests to internal endpoints and extract sensitive information. The post MLflow Vulnerability Exploited for Cloud Credential Theft appeared first on SecurityWeek .
AI Analysis
Technical Summary
CVE-2026-64849 is an unauthenticated server-side request forgery (SSRF) vulnerability in MLflow's Tracking Server. The vulnerability arises because the model-registry webhooks API is exposed without authentication, allowing attackers to send HTTP requests to internal endpoints. An SSRF protection introduced in version 3.10.0 can be bypassed, enabling attackers to reach cloud metadata services and exfiltrate cloud credentials and secrets. Exploitation began shortly after CVE assignment, primarily targeting cloud-hosted MLflow instances. The US Cybersecurity and Infrastructure Security Agency (CISA) has included this vulnerability in its Known Exploited Vulnerabilities catalog, recommending urgent patching.
Potential Impact
Attackers can exploit this vulnerability to send unauthorized HTTP requests to internal endpoints, including cloud metadata services, resulting in the theft of sensitive information such as cloud credentials and secrets. This can lead to unauthorized access to cloud resources and potential further compromise of affected environments. The vulnerability is critical due to its unauthenticated nature and the sensitivity of the data exposed.
Mitigation Recommendations
An official fix is available. All MLflow versions before 3.15.0 are affected, and upgrading to version 3.15.0 or later mitigates the vulnerability. Organizations running MLflow should prioritize patching exposed systems immediately. Additionally, review audit logs for signs of compromise and verify whether sensitive credentials have been exposed. Follow CISA's guidance for remediation timelines.
MLflow Vulnerability Exploited for Cloud Credential Theft
Description
The critical-severity flaw allows attackers to send HTTP requests to internal endpoints and extract sensitive information. The post MLflow Vulnerability Exploited for Cloud Credential Theft appeared first on SecurityWeek .
Affected software
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-64849 is an unauthenticated server-side request forgery (SSRF) vulnerability in MLflow's Tracking Server. The vulnerability arises because the model-registry webhooks API is exposed without authentication, allowing attackers to send HTTP requests to internal endpoints. An SSRF protection introduced in version 3.10.0 can be bypassed, enabling attackers to reach cloud metadata services and exfiltrate cloud credentials and secrets. Exploitation began shortly after CVE assignment, primarily targeting cloud-hosted MLflow instances. The US Cybersecurity and Infrastructure Security Agency (CISA) has included this vulnerability in its Known Exploited Vulnerabilities catalog, recommending urgent patching.
Potential Impact
Attackers can exploit this vulnerability to send unauthorized HTTP requests to internal endpoints, including cloud metadata services, resulting in the theft of sensitive information such as cloud credentials and secrets. This can lead to unauthorized access to cloud resources and potential further compromise of affected environments. The vulnerability is critical due to its unauthenticated nature and the sensitivity of the data exposed.
Mitigation Recommendations
An official fix is available. All MLflow versions before 3.15.0 are affected, and upgrading to version 3.15.0 or later mitigates the vulnerability. Organizations running MLflow should prioritize patching exposed systems immediately. Additionally, review audit logs for signs of compromise and verify whether sensitive credentials have been exposed. Follow CISA's guidance for remediation timelines.
Technical Details
- Classification
- {"confidence":0.77,"severitySource":"stated","classifier":"rss-v2"}
- Article Source
- {"url":"https://www.securityweek.com/mlflow-vulnerability-exploited-for-cloud-credential-theft/","fetched":true,"fetchedAt":"2026-08-20T17:06:20.840Z","wordCount":979}
Threat ID: 6a87340eacd9273b49e5d29e
Added to database: 08/20/2026, 17:06:22 UTC
Last enriched: 08/20/2026, 17:06:47 UTC
Last updated: 08/20/2026, 22:16:37 UTC
Views: 9
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
External Links
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.