South Australia's first conviction for sexual deepfakes under the new federal law: no jail time, and AI as a cheap tool for abuse

🕒 Published on Zendoric: October 29, 2026 · 00:15
✨ AI-generated · how it's made
William Yeates, 20, has been convicted of distributing sexual deepfakes of a young woman who was a minor when the harassment began. His prison sentence was suspended. The case shows that generative AI's harms are already everyday and close to home, not futuristic.
By News.com.au · 28 October 2026.
William Hamish Yeates, a 20-year-old law student at the University of Adelaide, has been sentenced in the South Australian District Court for creating and distributing fake sexual images and videos of a young woman he knew. He is the first person in the state to be convicted under the new federal deepfake offences (synthetic content that places a real person's face onto another body), News.com.au reports.
Judge Anne Barnett sentenced him to two years and three months in prison but suspended the term through a recognisance release order. Yeates will not go to jail. He had pleaded guilty to two counts of using a carriage service to harass and two counts of creating or altering sexual material without consent.
According to the court, he posted the material on X and on the pornographic site Ero Me over about five months, between August 2024 and February 2025. He even did so on Christmas Day. The victim was 17 when the offending began. The judge said the young woman first saw her graduation photos on an adult website, and that she now doubts her own judgement and whether she can trust others.
The ruling also sets out the context described by the judge: Yeates had become immersed in online communities centred on pornography, misogyny and Andrew Tate, and was seeking approval from those groups. Barnett found that he had not acted out of malice, that he had been desensitised by content he had consumed from a young age, and that he had shown genuine remorse. At an earlier hearing he read out a letter of apology in which he described his conduct as "grotesque".
**Our take.** What matters most in this case is not the tool but how little it takes to use it. A few years ago, producing a convincing fake sex video required technical skills. Today all it takes is a photo and an app that any student can get hold of. Broadly, that drop in cost is what turns deepfakes into a mass problem rather than a rarity: abuse no longer depends on the perpetrator's expertise.
The sentence also highlights a tension that will keep recurring. The law already exists and is being enforced, and that is a good sign: evidence-based governance can start by effectively prosecuting concrete, proven harms, without waiting for sweeping regulatory frameworks. But a suspended sentence, however understandable for a young and remorseful offender, is unlikely to deter others. And for the victim, material that has already been published does not disappear with a court ruling.
That is why the response cannot be purely criminal. The platforms that host and distribute this content, in this case X and a porn site, are the link that allows the harm to spread, and that is where the most can be done: detection, rapid takedown and traceability. Content provenance systems (watermarks and metadata indicating that something is synthetic) help, but only if they are mandatory and verifiable, and offenders can bypass them using open tools.
This is the hard side of the short term: an extraordinary technology that, while rules and culture catch up, also makes cruelty among young people cheaper. Acknowledging that is not doom-mongering. In the long run, the same generative capabilities will power medicine and abundance, but that will not happen by inertia. It requires laws that are enforced, platforms that respond, and education that makes clear that what is generated with a click has real victims.
🔗 Related on Zendoric
- A Tennessee janitor's case shows how generative AI has become a tool for creating child abuse · 2026-07-24
- Deepfakes in the office: the law already requires investigating harassment even when the evidence may be fake · 2026-09-09
- An AI-manipulated image, an email and an arrest: everyday deepfakes are now a school crime · 2026-10-22


