Deepfakes: the online abuse Cornwall parents are missing

Ask your child what a deepfake is, and you might get a blank look. Ask yourself the same question, and the odds of a confident answer aren’t much better either.

New findings show that almost two-thirds of children, and half of all parents, don’t know or understand the word, even as AI-generated abuse in the form of deepfakes becomes one of the fastest-growing safeguarding risks children face online.

What is a deepfake, really?

Dr Richard Anderson, Learning Experience Director at High Speed Training, describes a deepfake as a fake image, video or piece of audio made by manipulating something real. An ordinary holiday photo of someone with their family, for example. The manipulation can be convincing enough that the result becomes almost impossible to tell from the genuine thing.

Not every deepfake is made with bad intentions, and Dr Anderson is clear that they aren’t automatically illegal either. What has changed is how often they are now being used for harm. And when that happens, it can be a criminal offence.

“When abuse is not understood, it makes it difficult to recognise and teach children how to protect themselves and others,” Dr Anderson said. “A basic understanding of deepfakes and the safeguarding risks they can pose is essential to ensuring that children are kept safe and risks are identified and managed in a timely manner.”

Research by the online safety organisation Internet Matters found that a quarter of vulnerable children have had experiences with deepfakes depicting nudity, and that the great majority of deepfakes being made online are for harmful purposes. Nudity accounts for around 98% of all deepfakes in circulation, the research found, though the true scale of the problem is hard to pin down and could run higher still.

When does a deepfake break the law?

Deepfakes sit in a grey area for much of the time. But a growing list of uses tips them over into criminal territory. It includes:

• Intimate image abuse, sometimes called “revenge porn”
• Child sexual abuse material
• Hate crime
• Fraud
• False communications made to cause serious emotional or physical harm
• Terrorist activity
• Stalking and harassment
• Blackmail

For children specifically, the danger is being shown age-inappropriate content of people they know, or of themselves, without anyone’s consent. “Deepfakes violate an individual’s privacy, perpetuate harmful stereotypes and, at times, can be used to incite violence,” Dr Anderson said.

He added that the technology behind deepfakes is now easy enough for young people themselves to access, opening the door to child-on-child abuse, the spread of harmful opinions and an unsafe atmosphere in some schools. “Adults should therefore have an awareness of the risks that deepfakes pose and know what to do should a deepfake incident occur,” he said.

Three ways High Speed Training says parents can respond

The first is preventative education. Giving children a strong, early sense of what counts as bigotry or harmful content builds what Dr Anderson calls a zero-tolerance mindset. Children who understand the harm caused by hateful or violent content are less likely to engage with it themselves, and less likely to pass it on to others.

The second is talking about it openly. Engaging with the news around AI-generated abuse in an age-appropriate way, and hearing from survivors, can encourage children to pause and reflect on their own choices online. It isn’t about frightening anyone. Open conversation means children feel able to ask questions without fear of judgement, and it gives adults a better chance of spotting the early signs that something has happened, or might.

The third is critical thinking. For a generation that has never known life without the internet, whatever appears on a screen can carry a false sense of authority. Teaching children a healthy dose of scepticism, to question what they see rather than accept it automatically, helps them recognise false content before they engage with it or share it further.

High Speed Training has more detail in its own guide to AI-generated abuse and deepfakes.

Dr Anderson’s message for parents and schools is not to let the unfamiliarity of the word get in the way. “Education is a vital way for adults to inform young people of the danger of deepfakes,” he said. “Discussions around deepfakes must always be had in an age-appropriate manner, but it is vital that all young people, regardless of age, understand that deepfakes can be used to cause serious harm.”