We investigate how low-quality AI advisors, lacking quality disclosures, can help spread text-
based lies while seeming to help people detect lies. Participants in our experiment discern …