Building Trust in AI with Fiddler Auditor

The chapter discusses how Fiddler AI aims to instill trust in AI by introducing Fiddler Auditor, an open source tool designed to evaluate the robustness of large language models and ensure model accuracy, prevent biases, and protect against private data exposure. Through concepts like counterfactuals, the guest explains how Fiddler Auditor functions as a human questioner to test the AI's responses and uses feedback controls to detect issues like unsafe content and biases.