Aggregate Results

Live data from all completed evaluation sessions. Every AI injection was deliberately incorrect — a high average trust score is evidence of measurable automation bias across participants.