CVE-2025-33253
Published: 18 February 2026
Summary
CVE-2025-33253 is a high-severity Deserialization of Untrusted Data (CWE-502) vulnerability in Nvidia Nemo. Its CVSS base score is 7.8 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Exploitation for Client Execution (T1203); ranked at the 31.4th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog.
The strongest mitigations our analysis identified are NIST 800-53 SI-10 (Information Input Validation) and SI-2 (Flaw Remediation).
Threat & Defense at a Glance
Threat & Defense Details
Mitigating Controls (NIST 800-53 r5)AI
Applying vendor-provided patches for the deserialization flaw in NVIDIA NeMo Framework directly prevents exploitation by maliciously crafted files.
Validating the content and structure of files before deserialization in the framework blocks processing of untrusted data leading to RCE.
Verifying the integrity of framework components and loaded files using cryptographic mechanisms detects and prevents tampering or malicious deserialization payloads.
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
Deserialization of untrusted data in a crafted file directly enables client-side code execution (T1203) after a user loads the malicious file (T1204.002).
NVD Description
NVIDIA NeMo Framework contains a vulnerability where an attacker could cause remote code execution by convincing a user to load a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, denial of service, information disclosure,…
more
and data tampering.
Deeper analysisAI
CVE-2025-33253 is a vulnerability in the NVIDIA NeMo Framework, stemming from CWE-502 (Deserialization of Untrusted Data). It allows an attacker to potentially achieve remote code execution by convincing a user to load a maliciously crafted file. The issue carries a CVSS v3.1 base score of 7.8 (High), with vector AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H, indicating high impacts on confidentiality, integrity, and availability.
An attacker with local access and low privileges can exploit this vulnerability after tricking a user into loading the malicious file, leading to arbitrary code execution, denial of service, information disclosure, or data tampering. The low attack complexity and lack of required user interaction beyond the initial file load make it feasible for targeted exploitation in environments using the affected framework.
For mitigation details, security practitioners should consult official advisories, including the NVIDIA security bulletin at https://nvidia.custhelp.com/app/answers/detail/a_id/5762, the NVD entry at https://nvd.nist.gov/vuln/detail/CVE-2025-33253, and the CVE record at https://www.cve.org/CVERecord?id=CVE-2025-33253, which provide patch information and workarounds.
As part of NVIDIA's NeMo Framework for developing generative AI models, this vulnerability holds relevance for AI/ML practitioners securing training and inference pipelines against deserialization attacks. No public evidence of real-world exploitation has been reported.
Details
- CWE(s)