transparency

Making AI systems understandable: disclosing when AI is used, explaining decisions, and documenting model limitations, training data, and failure modes.

Syntax

ai-ethics

explainability: why did model predict X?
transparency: is this AI? what are its limits?

Example

ai-ethics

// Transparency practices:

// 1. Disclose AI use — tell users when AI is making decisions
"This credit decision was made by an automated system.
You can request a human review."

// 2. Explainability — provide reasons for decisions:
// LIME/SHAP for feature importance
import shap
explainer = shap.TreeExplainer(model)
shap_values = explainer.shap_values(X_test)
shap.summary_plot(shap_values, X_test)

// 3. Model cards — document model capabilities and limitations