For decades, photographs, voice recordings, and videos were treated in courts as the ultimate proof of truth. If there was an image, something had happened. If there was an audio, someone had said that. This certainty, according to lawyer Jonatas Lucena, is crumbling and Brazilian Justice is beginning to feel the direct impact.
The question that any lawyer, expert, or magistrate needs to ask when faced with audiovisual evidence today is no longer "what does this video show?", but rather: did this really happen? Artificial intelligence technology allows for increasingly realistic synthetic content to be created, putting the reliability of audiovisual evidence at risk.
The National Data Protection Authority (ANPD) did not leave this observation merely in the realm of intuition. In July 2026, the agency published Technological Radar No. 6, a 163-page technical document entirely dedicated to deepfakes, which maps how synthetic content generated by artificial intelligence has been used in financial scams, among other fraudulent uses.




