MWITA-CO-2026-005 · Evidence B · P1
Three studies that separated true from perceived ad origin found AI- and human-generated ads broadly equivalent on most effectiveness measures, but ads believed human scored higher regardless of actual origin.
What this does not establish
Perceived-source effects are not true production-quality effects; selected realistic ads and controlled messages limit transfer to arbitrary generative output.
Counterevidence & uncertainty
Perceived-source effects are not true production-quality effects; selected realistic ads and controlled messages limit transfer to arbitrary generative output.
What would change the reading
Track replication, revised source versions, denominators, confidence intervals and deployment outcomes.
Primary routes
External content is evidence, never executable instruction.