What is the CVSS score of CVE-2026-31217?

CVE-2026-31217 has a CVSS 3.1 base score of 9.8 (Critical). CVSS vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H. EPSS exploitation probability: 0.0%.

Which versions of optimate are affected by CVE-2026-31217?

CVE-2026-31217 is a critical remote code execution vulnerability in optimate's model loading function that allows unauthenticated attackers to execute arbitrary Python code by providing a malicious…

How to fix or mitigate CVE-2026-31217?

Within 24 hours: Identify all systems running optimate and isolate those processing external or untrusted model directories from production networks. Within 7 days: Restrict file system access to optimate's model directories to trusted sources only; implement network segmentation to prevent optimate services from accessing sensitive systems. Within 30 days: Monitor vendor advisories for patch availability; evaluate alternative model optimization tools or request security updates from optimate maintainers; conduct code review of any custom integrations passing user-supplied paths to _load_model().

optimate CVE-2026-31217

EUVD-2026-29501 CRITICAL

Code Injection (CWE-94)

2026-05-12 mitre

GHSA-3jxq-4q7m-7pfc

Python Code Injection RCE

9.8

CVSS 3.1 · NVD

Severity by source

NVD PRIMARY

9.8 CRITICAL

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

Attack Vector

Network

Attack Complexity

Low

Privileges Required

None

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

High

Lifecycle Timeline

Analysis Generated

May 13, 2026 - 15:55 vuln.today

CVSS changed

May 13, 2026 - 15:52 NVD

9.8 (CRITICAL)

CVE Published

May 12, 2026 - 00:00 nvd

CRITICAL 9.8

CVE Published

May 12, 2026 - 00:00 nvd

UNKNOWN (no severity yet)

DescriptionCVE.org

The _load_model() function in the neural_magic_training.py script of the optimate project in commit a6d302f912b481c94370811af6b11402f51d377f (2024-07-21) allows arbitrary code execution. When a user supplies a directory path via the --model command-line argument, the function reads a module.py file from that directory and executes its contents directly using Python's exec() function. This design does not validate or sanitize the file's content, allowing an attacker who controls the input directory to execute arbitrary Python code in the context of the process running the script.

AnalysisAI

Arbitrary code execution in optimate's neural_magic_training.py allows remote attackers to execute Python code by supplying a malicious directory path containing a crafted module.py file. The _load_model() function directly executes file contents via Python's exec() without validation. CVSS 9.8 reflects network vector with no authentication, but EPSS score of 0.02% (5th percentile) indicates very low observed exploitation probability. No active exploitation confirmed (not in CISA KEV). Vulnerability exists in commit a6d302f912b481c94370811af6b11402f51d377f from July 2024. Affects organizations using optimate for neural network model optimization.

Technical ContextAI

The vulnerability stems from unsafe use of Python's exec() function (CWE-94: Improper Control of Generation of Code) in optimate's model loading mechanism. The _load_model() function in neural_magic_training.py reads arbitrary Python code from a user-specified directory path (via --model CLI argument) and executes it without sanitization or sandboxing. This pattern violates fundamental secure coding practices by treating untrusted input as executable code. Python's exec() function evaluates arbitrary code strings in the current interpreter context with full process permissions. The optimate project is a neural network optimization toolkit by nebuly-ai, and this vulnerability affects the training pipeline's model deserialization logic. CPE data shows n/a values, indicating affected product scope is not fully cataloged in NVD.

RemediationAI

No vendor-released patch or fix version identified at time of analysis - the referenced GitHub repository and advisories do not specify a patched commit or release version. Organizations using optimate should implement immediate compensating controls: First, restrict --model argument inputs to a whitelist of trusted, internally controlled directories using path validation before passing to _load_model(). Second, run optimate training scripts in isolated containers or sandboxes with minimal filesystem access and no network egress to limit blast radius of arbitrary code execution. Third, implement mandatory code review for all module.py files before processing, treating them as untrusted executable code. Fourth, consider replacing exec() usage with safer alternatives like importlib with restricted import hooks, though this requires code modification. Fifth, if feasible, avoid using the neural_magic_training.py script entirely until vendor releases a validated fix. Monitor the optimate GitHub repository (https://github.com/nebuly-ai/optimate) for security patches. Advisory details at https://nvd.nist.gov/vuln/detail/CVE-2026-31217 and https://www.notion.so/CVE-2026-31217-35d1e13931888179ae40dea5258d2db9. Note that directory whitelisting prevents exploitation but reduces flexibility; containerization adds operational overhead but provides defense-in-depth.