🤖
AI Data Privacy
Privacy
Privacy controls and governance for data used to train, fine-tune, and run AI systems.
Definition
AI data privacy covers collection, labeling, training, fine-tuning, inference logs, and downstream sharing. It includes consent, minimization, retention, access control, and safeguards against re-identification and unintended memorization.
In plain English
Privacy controls and governance for data used to train, fine-tune, and run AI systems.
Why this matters
Why it matters: Training and logging can leak sensitive data and enable new forms of profiling.
Example
Example: Use short retention for prompts, restrict training on user content, and apply privacy-preserving aggregation where appropriate.