We’re attempting to provide an extensive analysis of the psychological and societal impacts of deepfakes and mass-produced AI-generated synthetic media, termed “AI slop.” We illustrate the immediate risks through cases like the fake Joe Biden robocall and a deepfake audio scandal that framed a school principal, underscoring how manipulated media can cause real-world harm even after being debunked. A major focus is on the continued influence effect and cognitive vulnerabilities, such as visual vividness and source amnesia, explaining why fake images and videos affect memory and belief long after their falsehood is revealed.
Furthermore, we explore the larger danger of the liar’s dividend, where the existence of fakes allows bad actors to dismiss legitimate evidence as false, leading to widespread epistemic confusion and a crucial erosion of democratic trust.
