Publications

Research 04.2026

Confidence Without Calibration

Five AI models, fifteen verifiable questions, and no uncertainty when the answers were wrong.

Research 03.2026

The Label Says Otherwise

What eight AI platforms get wrong about GLP-1 medications.

Research 02.2026

Houston, We Have a Problem

Documented cases of AI platforms generating falsehoods about real organizations.

Perspective 01.2026

A Letter from the Founder

On the transition from search-driven reputation to AI-driven information, and why this firm exists.