How to meet it
OWASP AISVS v1.01
Model Lifecycle Management & Change Control · Secure Development Practices
Keep training apart from production
Verify that model training and fine-tuning environments are isolated from production environments.
OWASP AISVS v1.01, requirement 3.4.2
In plain English
Run training and fine-tuning away from production, so a problem in training cannot reach live systems and live data is not exposed to experiments.
What good looks like
- Training runs in separate accounts or networks.
- Training jobs cannot write to production systems.
- Moving a model to production goes through the deployment pipeline.
How teams usually do it
- Separate cloud accounts with no direct network route to production.
- A promotion step that copies approved artifacts into production.
Evidence an auditor would accept
- Network and account design.
- Promotion pipeline configuration.
Common mistakes
- Training jobs with credentials that can also deploy to production.
Not sure which requirements apply to your system? Take the assessment.
Requirement text from the OWASP AI Security Verification Standard, licensed under CC BY-SA 4.0. This guidance is shared under the same licence.