by Dr Helena Ivanov
In the latest in a growing series of AI-created deepfakes, the newest victim was Italian Prime Minister Giorgia Meloni. A few days ago, Meloni posted fake images of herself, images that portrayed her in a sexualised position, wearing scanty underwear, and described deepfakes as a “dangerous tool” that can harm and target anyone. She made clear that while she is strong enough to confront and respond to this kind of content, many others may not be. She also urged people to verify content before sharing it or believing it.
Importantly, this is not the first time Meloni has been targeted by deepfakes. Nor is she alone. Many high-profile women have been victims of the same crime, with doctored and sexualised images of them appearing on pornographic websites. Ultimately, the Italian government passed a law criminalising deepfakes that cause “unjust harm” to the person depicted. That Italy is taking the problem seriously is also demonstrated by its decision to ban Clothoff, an app that “uses AI to generate fake nude images of real people.” Moreover, Meloni herself has taken legal action in the past, suing two men for €100,000 after they allegedly produced fake videos of the Prime Minister and posted them on a US pornographic website.
Italy is not an isolated case. Across the globe, female politicians are increasingly being targeted by AI-generated pornographic deepfakes. And, in fact, it is not only politicians who are being targeted. As HJS has previously written, Elon Musk and his AI chatbot Grok have come under fire in the UK and several EU countries, including France, after users were reportedly able to use Grok to create sexualised images of minors and women without their consent. Ofcom is currently investigating the incident, while the UK government is considering the criminalisation of both the creation and dissemination of deepfakes. In France, X’s offices have been raided, and the investigation into the matter continues.
However, while governments are still trying to grapple with the scale of the problem, more and more people are becoming victims of this vicious technology. There is simply no time for long inquiries that drag on indefinitely. While those inquiries are ongoing, victims continue to suffer serious harm on a daily basis. Any country that is genuinely committed to protecting its citizens, particularly, in this case, minors and women, must act quickly. Likewise, if the integrity of democratic processes is to be protected, so too must the female politicians who are disproportionately targeted by these technologies.
As we finalise our forthcoming report on AI and disinformation, this latest blog argues that this technology must be stopped. The creation of pornographic deepfakes does not, in any meaningful sense, constitute freedom of expression. Nor does it serve democratic norms. Instead, it is being used to viciously harm citizens in democratic societies.
For that reason, we advocate for the urgent ban and criminalisation of the creation and dissemination of pornographic deepfakes. We argue in favour of criminalising and penalising anyone who creates, distributes, or knowingly amplifies this content. Likewise, we urge governments to put pressure on companies to implement the necessary safeguards to prevent people from abusing AI technologies to create deepfakes. Where companies refuse to comply, and continue to allow the proliferation of these materials, we also argue in favour of penalising those companies.
There are many areas in which restricting disinformation is complicated and fraught, particularly where such restrictions risk undermining freedom of expression. But the creation of pornographic deepfakes is not one of those complex cases. In fact, it is very simple. It is abuse. It is harm. And it must be banned.
