What is a deepfake? Channel 4 and Vicky Pattison's controversy

She is a British television personality, best known for appearing on reality TV shows such as Geordie Shore and I'm a Celebrity...Get Me Out of Here! She is also a presenter, author, and podcast host.

Pattison wanted to use the new documentary to illustrate how adult content can spread online and how these images and videos are taken down.

However, online image abuse survivors, such as 23-year-old Jodie, advised against airing the video. She told the Guardian that the decision to use the deepfake footage of Pattison “lacked compassion” and was in “poor taste”.

“Survivor organisations strongly advised against it when they were approached by producers,” she said. “It’s just in really, really poor taste that they sought out advice and then ignored it completely.”

While the documentary may be controversial, it has raised important questions about deepfakes and their impact.

Deepfakes are often associated with pornography, but they are not limited to adult content or videos. While pornography is one of the most damaging forms of deepfake content, it is not the only harmful application.

So, how do deepfakes work?

A deepfake typically involves digitally altering a person’s face or body to make them appear as someone else. In the case of deepfake audio, a person’s voice is synthesised to create a clip that sounds like they are saying something they never actually did. These manipulations are powered by machine learning and artificial intelligence.

A 2019 Deeptrack Labs study found that 96 per cent of deepfake videos were pornographic. However, less explicit deepfake content on social media has surged since then, as the tools to create them have become more widely accessible.

For example, you may have seen Miles Fisher’s online content. He posts deepfake videos of Tom Cruise doing humorous things on TikTok. Fisher’s voice and real body are seen but Cruise’s face is digitally placed over his.

It certainly helps that Fisher’s appearance and voice are strikingly similar to a younger Cruise.

Deepfake tech is most dangerous when it is used to create pornographic content without people’s consent, or when it is used as a tool to spread misinformation or disinformation.

Even the most ostensibly harmless deepfake content presses some of the same buttons as the malicious stuff, nudging your brain to believe it’s real even if it is flagged as deepfake.

A deepfake video might show a political candidate saying something incendiary before an election, in an attempt to hurt their chances of winning. It could be a public figure speaking out against a vaccine, despite never having done so.

Tools have been developed — and are continually being improved — to keep pace with the growing sophistication of deepfakes. However, due to the viral nature of some of this content, the damage is often done by the time videos or audio clips are exposed as fakes.

In deepfake porn, the face of a celebrity or public figure is transplanted into a sex scene, effectively turning any piece of porn into the kind of “sex tape” content that used to attract so much attention years ago.

Actresses Margot Robbie and Scarlett Johansson are among those who have been victims of deepfake pornography.

“I think it’s a useless pursuit, legally, mostly because the internet is a vast wormhole of darkness that eats itself. It’s a fruitless pursuit for me but a different situation than someone who loses a job over their image being used like that.”

She said: “The internet is just another place where sex sells and vulnerable people are preyed upon. And any low-level hacker can steal a password and steal an identity. It’s just a matter of time before any one person is targeted.”

Deepfakes, in general, are not illegal in the UK, as they cover a wide spectrum — from innocent face-swap apps to harmful or malicious uses, such as pornographic videos involving public figures. Banning deepfakes outright is not practical, given the range of applications and technological innovation behind them.

However, the UK Government has taken action to address harmful uses of deepfake technology. In November 2022, it announced plans to criminalise the sharing of pornographic deepfakes that depict individuals without their consent. These explicit, manipulated images or videos would be covered by an amendment to the Online Safety Bill, which aims to enhance online protections and tackle non-consensual intimate content.

Beyond this, other malicious uses of deepfakes could fall under existing laws, such as fraud, defamation, or libel. For example, if a deepfake is used to deceive or defraud someone or damage their reputation, it could lead to criminal or civil action.

However, as of January 2025, there is no blanket legal protection against someone using your likeness in a deepfake, meaning individuals are not automatically covered if their image is used this way — unless it falls under one of the specific categories of harm addressed by the law.

The UK’s legal landscape around deepfakes continues to evolve. However, the focus for now is on tackling the most harmful and malicious uses while balancing technological innovation with protecting individual rights.