Deepfakes and the Law

What if someone uses AI to create a fake version of your voice for their own aims? Recently, the actor, broadcaster and writer Stephen Fry found that someone had recreated his voice to narrate a documentary without his knowledge. What does the law have to say about deepfakes? What are your rights, and in which circumstances could someone be sued, or prosecuted? Associate solicitor Oliver Lock of Farrer & Co explains what the law can, and can't do.

Creating fakes with AI, and the software to detect them, is a growing field. The same is true of forensic speech recognition, which is done both by ear and machine and can help the police or a court identify whether a recording is the voice of a suspect, for example. Dr Anil Alexander of Oxford Wave Research Ltd plays some samples to presenter Joshua Rozenberg. Can he guess them right? And what other uses are there for this technology in law enforcement?

Forensic scientists are often called upon to give evidence in court, as are doctors. These expert witnesses are crucial, but things can go wrong. Some find cross-examination so bruising that they don't want to repeat it. Others fear for their reputation, if they're pushed into saying something they hadn't meant to say. Baroness Professor Sue Black is a leading forensic anthropologist and shares her thoughts.

Sometimes barristers and judges are out of their depth on the science of a case. One solution to this problem has been put forward by the independent scientific academy the Royal Society, with the Royal Society of Edinburgh: subject-specific primers on relevant topics. As Dame Dr Julie Maxton, executive director of the Royal Society explains, leading scientists write and peer-review the primers, such as on ballistics or DNA, and senior judges cross-check them from the legal perspective. The primers are online, aimed at judges but available for everyone. The hope is that if barristers fail to ask the right questions on the science, judges who've read the primers can then do so instead.

Presenter: Joshua Rozenberg Producer: Arlene Gregorius Researcher: Diane Richardson Editor: Clare Fordham Sound engineers: James Beard and Rod Farquhar Production Coordinators: Maria Ogundele and Rosie Strawbridge