During an AI model audit, the auditor requests access to the model's training data but is told it was deleted after training to reduce storage costs. What risk does this create for ongoing model governance?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — deleting training data creates multiple governance failures: training cannot be reproduced or audited, bias investigations cannot trace root causes to specific data subsets, and data subject rights cannot be fully honored. Missing training data is not an inference cache, so latency is not the issue.
Full Explanation
Inability to reproduce model training, investigate bias claims, or respond to data subject access requests related to training data is correct because deleting training data blocks reproducibility, bias root-cause analysis, and verification of data subject rights. Increased inference latency due to missing training data cache is wrong because training data is not cached for inference. Regulatory penalty for violating data retention minimums under GDPR is wrong because GDPR emphasizes minimization, not mandatory indefinite retention. Reduced model accuracy due to absence of training data for retraining is wrong because that is a practical retraining limitation, not the primary governance risk.