How to meet it
OWASP AISVS v1.01
Training Data Integrity & Traceability · Training Data Origin & Data Security
Keep an inventory of training data sources
Verify that an up-to-date inventory is kept of every training-data source, including its origin, responsible party, license, collection method, intended use constraints, and processing history.
OWASP AISVS v1.01, requirement 1.1.2
In plain English
Keep a current list of every source your training data comes from: where it came from, who owns it, the licence, how it was collected, what it may be used for and what has been done to it since.
What good looks like
- Every dataset used in training appears in the inventory, including scraped, bought and synthetic data.
- Each entry names an owner who can answer questions about it.
- Licence and use restrictions are recorded and checked before a dataset is used for a new model.
How teams usually do it
- A data catalogue entry per source, using a standard template such as datasheets for datasets.
- A pipeline check that refuses to train on a dataset with no catalogue entry.
Evidence an auditor would accept
- The inventory itself, with owners and licences filled in.
- A recent training run's dataset list, matched against the inventory.
Common mistakes
- An inventory written once for the first model and never updated.
- Leaving out data pulled in by a fine-tuning or evaluation step because it was "only a small set".
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.