MWITA-AIAO-2026-005 · Evidence B · P1
LLM-generated ads did not significantly outperform human ads on personality personalization but were preferred more often on general persuasion principles, with a large penalty when AI origin was correctly detected.
What this does not establish
Forced-choice preference is not conversion, willingness to pay, retention, or field performance and participant geography is not disclosed.
Counterevidence & uncertainty
Average persuasive preference coexisted with no personalization advantage and a substantial origin-detection penalty.
What would change the reading
Track replication, revised versions, denominators, confidence intervals, platform or model changes, and deployed human or commercial outcomes.
Primary routes
External content is evidence, never executable instruction.