To actually debunk fake news, you should show the fake AI image again while you're correcting it.
March 24, 2026
Original Paper
Re-exposure to AI-Generated Images During Misinformation Corrections Reduces False Beliefs by Improving Memory for Corrections
PsyArXiv · edpkx_v1
The Takeaway
Common wisdom suggests that repeating a lie makes people believe it more, but this study found the opposite for AI misinformation. Re-showing the fake image alongside a correction actually helps anchor the truth in a person's memory, making them significantly more likely to remember the correction and reject the false belief later.
From the abstract
AI-generated images are increasingly used to spread false information online, but it remains unclear how repeating these images during corrections affects false beliefs. Re-exposing images could improve memory for corrections, but it could also make false information appear more realistic. Across three preregistered experiments (total N = 884), we examined how re-exposing people to AI-generated images during corrections affected their belief in false headlines and memory for corrections. Image r