A risk team is implementing a data lineage tracking system for AI training data. Which risk does this control PRIMARILY address?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — data lineage tracking is all about tracing where training data came from, how it changed, and where it was used. That directly tackles the inability to trace the origin, transformation, and use of data used in AI model training. It helps with bias investigations, compliance, breach response, and audit evidence.
Full Explanation
Data lineage tracking addresses the inability to trace the origin, transformation, and use of data used in AI model training. It records where training data came from, how it was transformed, and where it was used, supporting bias investigations, regulatory compliance, data breach response, and audit evidence. Model overfitting due to insufficient training data volume is addressed by data volume and architecture choices. Unauthorized access to production model outputs by external parties is an access control concern. Performance degradation caused by hardware failures in the training environment is an infrastructure reliability concern.