COMMON WEAKNESS ENUMERATIONDraft

CWE-1434

Insecure Setting of Generative AI/ML Model Inference Parameters

The product has a component that relies on a generative AI/ML model configured with inference parameters that produce an unacceptably high rate of erroneous or unexpected outputs.

Abstraction

Base

Structure

Simple

Likelihood of Exploit

Medium

Understand the problem

What is the problem?

The product has a component that relies on a generative AI/ML model configured with inference parameters that produce an unacceptably high rate of erroneous or unexpected outputs.

Extended description

Generative AI/ML models, such as those used for text generation, image synthesis, and other creative tasks, rely on inference parameters that control model behavior, such as temperature, Top P, and Top K. These parameters affect the model's internal decision-making processes, learning rate, and probability distributions. Incorrect settings can lead to unusual behavior such as text "hallucinations," unrealistic images, or failure to converge during training. The impact of such misconfigurations can compromise the integrity of the application. If the results are used in security-critical operations or decisions, then this could violate the intended security policy, i.e., introduce a vulnerability.

Root cause

How and when does the problem arise?

Lifecycle phase

Build and Compilation

During model training, hyperparameters may be set without adequate validation or understanding of their impact.

Lifecycle phase

Installation

During deployment, model parameters may be adjusted to optimize performance without comprehensive testing.

Lifecycle phase

Patching and Maintenance

Updates or modifications may be made to the model that alter its behavior without thorough re-evaluation.

Risk

What does successful exploitation lead to?

Varies by ContextUnexpected State

Affected security scopes: Integrity, Other

The product can generate inaccurate, misleading, or nonsensical information.
Alter Execution LogicUnexpected StateVaries by Context

Affected security scopes: Other

If outputs are used in critical decision-making processes, errors could be propagated to other systems or components.

Defence

How is it prevented and fixed?

Implementation · System Configuration · Operation

Defence to apply across several lifecycle phases

Detail
Develop and adhere to robust parameter tuning processes that include extensive testing and validation.

Implementation · System Configuration · Operation

Defence to apply across several lifecycle phases

Detail
Implement feedback mechanisms to continuously assess and adjust model performance.

Documentation

Documentation aşaması savunması

Detail
Provide comprehensive documentation and guidelines for parameter settings to ensure consistent and accurate model behavior.

Verification

How is it detected?

Automated Dynamic Analysis

Moderate
Manipulate inference parameters and perform comparative evaluation to assess the impact of selected values. Build a suite of systems using targeted tools that detect problems such as prompt injection (CWE-1427) and other problems. Consider statistically measuring token distribution to see if it is consistent with expected results.

Given the large variety of outcomes, it can be difficult to design testing to be comprehensive enough, and there is still a risk of unpredictable behavior.

Manual Dynamic Analysis

Moderate
Manipulate inference parameters and perform comparative evaluation to assess the impact of selected values. Build a suite of systems using targeted tools that detect problems such as prompt injection (CWE-1427) and other problems. Consider statistically measuring token distribution to see if it is consistent with expected results.

Given the large variety of outcomes, it can be difficult to design testing to be comprehensive enough, and there is still a risk of unpredictable behavior.