Skip to content
News & Insights
News & Insights
Evaluate the Model, Not the Brand
Two students, seven AI systems, and why a brand name does not predict equitable behavior.
How Long Is Long Enough? Conversation Length in AI Safety Testing
Why a five-message safety test can miss what a longer conversation reveals.
ICC: Grading the Graders
Why every score needs a reliability figure, and where agreement stops.
Working with BU Wheelock to strengthen AI evaluation in education
Combining our strengths with BU Wheelock to advance the K-12 Education Index.
Methodology
AI Scorecards
AI Scorecards
Equity in K-12 Education
Four leading models scored against six equity indicators.
Education
K–12 & Education
Education Overview
Protecting students, empowering educators, building trustworthy EdTech.
Independent product evaluation
Private, collaborative testing for AI products used with K-12 students.
League of AI-Fluent Schools
A national learning community for evidence-driven AI in schools.
Request a briefing
Hello world!
Welcome to WordPress. This is your first post. Edit or delete it, then start writing!
