What is the CVSS score of CVE-2026-24150?

CVE-2026-24150 has a CVSS 3.1 base score of 7.8 (High). CVSS vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H. EPSS exploitation probability: 0.1%.

Which versions of Nvidia are affected by CVE-2026-24150?

NVIDIA Megatron-LM contains a critical remote code execution vulnerability in checkpoint file loading that could allow attackers to execute arbitrary code if users load malicious model files.

How to fix or mitigate CVE-2026-24150?

Within 24 hours: Inventory all systems running Megatron-LM and restrict checkpoint file loading to trusted internal sources only; disable remote checkpoint loading if not operationally critical. Within 7 days: Implement file integrity monitoring on checkpoint repositories and conduct security awareness training on social engineering risks related to model file distribution. Within 30 days: Establish a secure checkpoint validation pipeline, monitor NVIDIA security advisories for patch availability, and prepare for rapid deployment once patches are released.

Nvidia CVE-2026-24150

EUVD-2026-15005 HIGH

Deserialization of Untrusted Data (CWE-502)

2026-03-24 nvidia

GHSA-r99w-qgv6-8r5p

RCE Information Disclosure Nvidia Deserialization

7.8

CVSS 3.1 · NVD

Severity by source

NVD PRIMARY

7.8 HIGH

AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Primary rating from NVD · only source for this CVE.

CVSS VectorNVD

CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Attack Vector

Local

Attack Complexity

Low

Privileges Required

Low

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

High

Lifecycle Timeline

EUVD ID Assigned

Mar 24, 2026 - 20:31 euvd

EUVD-2026-15005

Analysis Generated

Mar 24, 2026 - 20:31 vuln.today

CVE Published

Mar 24, 2026 - 20:24 nvd

HIGH 7.8

DescriptionCVE.org

NVIDIA Megatron-LM contains a vulnerability in checkpoint loading where an Attacker may cause an RCE by convincing a user to load a maliciously crafted file. A successful exploit of this vulnerability may lead to code execution, escalation of privileges, information disclosure, and data tampering.

AnalysisAI

NVIDIA Megatron-LM contains an unsafe deserialization vulnerability (CWE-502) in its checkpoint loading functionality that allows remote code execution when a user is tricked into loading a maliciously crafted checkpoint file. The vulnerability affects NVIDIA Megatron-LM installations and can lead to code execution, privilege escalation, information disclosure, and data tampering with a CVSS score of 7.8. There is no current indication of active exploitation in CISA's KEV catalog, and EPSS data was not provided in the intelligence sources.

Technical ContextAI

NVIDIA Megatron-LM is a large-scale transformer model training framework used for developing large language models. The vulnerability stems from CWE-502 (Deserialization of Untrusted Data), where the checkpoint loading mechanism fails to properly validate serialized data before deserializing it. According to the CPE string (cpe:2.3:a:nvidia:megatron_lm:*:*:*:*:*:*:*:*), all versions of the product are potentially affected. Checkpoint files in machine learning frameworks typically contain model weights, optimizer states, and configuration data serialized using formats like pickle in Python, which can execute arbitrary code during deserialization if the source is untrusted. This class of vulnerability is particularly dangerous in ML frameworks because checkpoint files are frequently shared between researchers and systems.

RemediationAI

Consult the official NVIDIA security advisory at https://nvidia.custhelp.com/app/answers/detail/a_id/5769 for patched versions and upgrade instructions specific to your Megatron-LM deployment. Until patching is completed, implement strict controls around checkpoint file provenance by only loading checkpoints from verified and trusted sources, implementing file integrity verification using cryptographic signatures, and running checkpoint loading operations in sandboxed or containerized environments with minimal privileges. Organizations should audit their current checkpoint sources and establish policies prohibiting the use of checkpoints from public repositories or unverified third parties. Consider implementing allowlisting for checkpoint file sources and conducting security reviews of any shared model files before loading them into production or development systems.