Deepfakes pose challenge to criminal trials, says Senior Advocate CU Singh

Senior Advocate Chander Uday Singh has voiced significant concerns regarding the potential impact of AI-generated deepfake technology on criminal justice systems. Speaking at a recent discussion on AI in the legal profession, Singh highlighted the increasing sophistication of synthetic media.

Singh stated that the current level of "absolute photorealism and video realism" allows for the creation of convincing fake content, including fabricated audio and video. "A voice transcript which the person will not be able to deny, which no forensic expert will be able to deny against, that this was Chander Uday Singh’s voice, even though I've not said that ever," he explained. He emphasized that the time and expertise required to debunk such fabricated evidence could become a major bottleneck as the volume of AI-generated content grows.

Further elaborating on the risks, Singh warned that deepfakes could be used to fabricate circumstantial evidence, which often plays a crucial role in criminal trials when direct evidence is scarce. He expressed concern that "a wily and unscrupulous prosecutor could generate enough stuff to tie that person up in knots in court," particularly in jurisdictions where evidence tampering is already an issue. He also raised the point that defendants may lack the financial resources to effectively counter such sophisticated fabricated evidence.

Singh urged the judiciary and legal bodies to proactively develop safeguards against the misuse of this technology. "If you don’t get on the horse right now and start holding its bridle and controlling it, it’s going to be a runaway wreck," he cautioned, stressing the need for regulatory action before the problem becomes unmanageable.