CISA warns of hackers exploiting critical MLflow vulnerability
The Cybersecurity and Infrastructure Security Agency (CISA) warned federal agencies that threat actors are now exploiting a critical MLflow vulnerability. This vulnerability can be used by attackers without privileges to remotely access internal services or cloud metadata…
Intelligence analysis by Llama
CISA has warned federal agencies that threat actors are exploiting a critical MLflow vulnerability, which can be used to remotely access internal services or cloud metadata configurations on unpatched instances. This vulnerability was patched in version 3.15.0, but many instances may still be unpatched.
Imagine you have a super powerful computer that can do lots of things, but it's also very vulnerable to hackers. This is what's happening with the MLflow vulnerability. Hackers can use this vulnerability to get into the computer and steal important information. To fix this, we need to update the computer's software so that it's not vulnerable anymore.
Analysis
MLflow Vulnerability Overview
The MLflow vulnerability, tracked as CVE-2026-64849, is a critical DNS-rebinding server-side request forgery (SSRF) bypass in MLflow's outbound webhook delivery. This vulnerability can be used by attackers without privileges to remotely access internal services or cloud metadata configurations on unpatched instances.
Impact of the Vulnerability
Successful exploitation of this vulnerability can allow threat actors to steal cloud credentials, such as AWS Identity and Access Management (IAM) credentials, in low-complexity attacks. This can give attackers partial or total control of a targeted system.
CISA Warning
CISA has warned federal agencies that threat actors are now exploiting this vulnerability and has ordered U.S. Federal Civilian Executive Branch (FCEB) agencies to secure their MLflow instances within two weeks as mandated by Binding Operational Directive 26-04.
Ongoing Attacks
Cybersecurity firm watchTowr revealed that attackers began scanning for MLflow systems within hours after the CVE ID was assigned. Attackers are exploiting the vulnerability to reach cloud metadata services directly, and exfiltrating cloud credentials and secrets.
Prevention and Mitigation
To prevent exploitation of this vulnerability, organizations that run MLflow should prioritize patching any exposed systems, review audit logs for signs of compromise, and check whether sensitive credentials may have been exposed.
Key points
- CISA has warned federal agencies that threat actors are exploiting a critical MLflow vulnerability.
- This vulnerability can be used by attackers without privileges to remotely access internal services or cloud metadata configurations on unpatched instances.
- CISA has ordered U.S. Federal Civilian Executive Branch (FCEB) agencies to secure their MLflow instances within two weeks as mandated by Binding Operational Directive 26-04.
- Cybersecurity firm watchTowr revealed that attackers began scanning for MLflow systems within hours after the CVE ID was assigned.
If this vulnerability is patched quickly, it's possible that the number of attacks will decrease, and the risk of data breaches will be reduced. Additionally, the fact that CISA has warned federal agencies about this vulnerability may lead to increased awareness and vigilance among organizations that use MLflow.
If the vulnerability is not patched quickly, it's possible that the number of attacks will increase, and the risk of data breaches will be higher. Additionally, the fact that attackers are already exploiting this vulnerability may indicate that they have a significant advantage over organizations that are trying to protect themselves.
