Code Injection Vulnerability in NVIDIA Megatron-LM
CVE-2025-23306
7.8HIGH
What is CVE-2025-23306?
NVIDIA Megatron-LM exhibits a significant security vulnerability in the megatron/training/arguments.py component. An attacker can exploit this flaw by supplying malicious input, potentially allowing for unauthorized code execution. Successful exploitation can lead to escalation of privileges, sensitive information disclosure, and data tampering, posing severe risks to the integrity and confidentiality of the system. Immediate attention is recommended to mitigate these threats.
Affected Version(s)
Megatron-LM All platforms All versions prior to 0.12.2
References
CVSS V3.1
Score:
7.8
Severity:
HIGH
Confidentiality:
High
Integrity:
High
Availability:
High
Attack Vector:
Local
Attack Complexity:
Low
Privileges Required:
Low
User Interaction:
None
Scope:
Unchanged
Timeline
Vulnerability published
Vulnerability Reserved