The rapid advancements of generative AI (GenAI) technologies have brought the issue of data manipulation to a whole new level. Now individuals can easily create highly realistic fabricated media, including videos, audio, and images, that are increasingly difficult to detect. These developments raise broad societal concerns, particularly in legal contexts, where authenticity of evidence is central to ensuring fairness. This review links documented legal cases to psychological research on memory distortion and false confessions to explore how AI-generated evidence affects individuals’ perception and sense of self and the consequences in a legal context. Decades of research have shown that memory’s vulnerability to suggestion, especially highlighting the power of personalized misinformation. Even in the pre-AI era, studies have shown how false memories have been induced using suggestion and simple media manipulations and how these have led to false confessions. GenAI adds another level of concern, yet no research directly examines its potential to distort memory and to elicit false confessions. This paper identifies urgent questions: How does AI-generated evidence impact memory? How can individuals ensure authenticity of the evidence confronting them? And how can legal systems prepare for the erosion of trust brought upon by this uncertainty? Proposed solutions like watermarking and implementation of authenticity protocols offer partial relief; however, the rapid advancements make sure that no definitive safeguards can be implemented. Understanding how GenAI interacts with our memory vulnerabilities is critical not only for legal psychology but for safeguarding the integrity of justice in today’s digitally mediated society.Keywords: Generative AI, deepfakes, memory distortion, legal psychology, false confessions, trust in evidence
Iannuzzi et al. (Wed,) studied this question.