Core AI Concepts
Fine-tuning
Fine-tuning is additional training that adapts a base AI model to a specific task or organization using a curated dataset — a dataset that often contains the sensitive examples.
Fine-tuning continues a model's training on a smaller, targeted dataset — support transcripts, legal documents, code — so it performs better on a specific domain. Organizations fine-tune for tone, format, and expertise the base model lacks.
The governance question is the dataset: fine-tuning data frequently contains real records, and a model can memorize and regurgitate rare strings from it. De-identification before fine-tuning, and access controls on the resulting model, are the standard mitigations.
Where this shows up
Related terms
See it in your own organization.
Sanitized AI inventories the AI tools in use and redacts sensitive data from prompts before it leaves.