Code execution via unsanitized Hydra target metadata
Published Aug 13, 2025 · Updated Feb 26, 2026
Code injection in NVIDIA NeMo Framework before 2.3.2 allows local users to execute code through crafted .nemo model metadata. The restore_from and from_pretrained paths pass unsanitized model_config.yaml target values to hydra.utils.instantiate, which imports and instantiates attacker-selected objects. Low-privileged local access is required, and injected code runs with the loading process's permissions, enabling data tampering.
Summary
What happened
Code injection in NVIDIA NeMo Framework before 2.3.2 allows local users to execute code through crafted .nemo model metadata. The restore_from and from_pretrained paths pass unsanitized model_config.yaml target values to hydra.utils.instantiate, which imports and instantiates attacker-selected objects. Low-privileged local access is required, and injected code runs with the loading process's permissions, enabling data tampering.
The record
- CVE
- CVE-2025-23304
- Published
- Aug 13, 2025
- Updated
- Feb 26, 2026
- Vendor
- DPDK Project
- Product
- NVIDIA NeMo Framework
- Classifications
- CWE-22, T1203
- Attack vector
- local
- Privileges
- authenticated
Timeline
How it unfolded
- Aug 13, 2025CVE publishedPublication date reported by the CVE source.
- Feb 26, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=All versions prior to 2.3.2
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Unit 42 malicious NeMo model metadata exampleproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo