The rise of deepfake technologies is transforming the landscape of digital evidence in U.S. courts. AI-driven manipulation of audio, video, and photographic evidence is increasingly undermining the pursuit of justice, making it difficult for judges, juries, and forensic experts to distinguish between authentic and fabricated content. Challenges presented by these technologies have already surfaced in cases like United States v. Guy Wesley Reffitt, Sz Huang v. Tesla, Inc., and State of Washington v. Puloka. In this session, our panelists will explore how generative AI’s accessibility is enabling almost anyone to alter evidence and create convincing deepfake content, raising significant concerns for the administration of justice. They will address the concept of the "liar’s dividend";—the harmful advantage deepfakes provide by casting doubt on legitimate evidence. They will discuss the limitations of existing admissibility standards under the Daubert and Frye rules, and the potential but inconsistent use of Rule 403 as a gatekeeper for deepfake evidence. Additionally, they evaluate practical current approaches for screening, assessing, and ultimately admitting digital evidence. Join us as we consider this looming challenge to justice, the role of lawyers and judges in mitigating the risks, and the measures needed to safeguard the integrity of the justice system against the growing deepfake threat.