Deep fake and deep nude

AI is an ever-evolving technology that is having a significant impact on the way we live and interact with the world around us, bringing several innovations in many fields while at the same time bringing new challenges for privacy and personal data protection.

Among these challenges, deep fakes and deep nude is a topic of increasing prevalence and relevance, entailing a number of risks and critical issues. In this article we’ll explore the main issues and possible negative consequences of these technologies with particular reference to identity theft, cyberbullying and revenge porn, and the right to be forgotten.

 

What are deep fakes and deep nudes?

Deep fakes are videos or images created using AI to replace one person’s face with another’s. Often it’s so realistic that they are almost indistinguishable from reality. The word deep fake is a neologism created by merging the terms “fake” and “deep learning”, a particular AI technology. The starting point is always the real faces, real bodies, and real voices of people (all “personal data” under the GDPR), but transformed into digital “fakes”. Unfortunately, deep fakes are not created for purely goliardic purposes, but can also be used for illicit purposes, such as spreading fake news or compromising a person’s reputation.

Deep nudes, on the other hand, are a subcategory of deep fakes, consisting of images manipulated by AI for the purpose of removing a person’s clothing, creating fake, sexually explicit images. This technology has been used to make pornographic images of celebrities and other people without their consent, and it seems clear that deep nudes can also be used for extortion or defamatory purposes.

 

The main risks and critical issues: From identity theft to difficulties in exercising the right to be forgotten

The ease with which sensitive attributes of an individual’s personal sphere can be manipulated using technologies such as deep fakes is evident. The situation is further complicated by the increasing availability of apps and software that enable the creation of highly sophisticated deep fakes with common devices such as smartphones, which are now within the reach of everyone, including minors.

 

Identity theft

As also pointed out in the Vademecum of the Italian Data Protection Authority, the phenomenon of deep fakes can give rise to particularly serious forms of identity theft, as the people involved completely lose control over their own image and public representation of themselves.

Not only the physical representation of the people involved in the deep fake can be manipulated, but also the context in which they are placed, the people around them, and the situations they represent. Moreover, the possible negative consequences can extend beyond the victim’s personal sphere, for example, by damaging the reputation of a company or organization with which the person is associated.

 

Cyberbullying and revenge porn

Deep fake videos can be created with the intent to engage in actual acts of cyberbullying or even revenge porn, that is the online sharing – for the purpose of blackmail, denigration, or revenge, by former partners, lovers, or rejected admirers – of photos and videos with sexual or even pornographic content, which, in the case of deep nude, are obviously fake.

In the specific case of deep nudes, the faces of people (including minors) can be “grafted”, using special software, onto the bodies of other subjects, either naked or engaged in poses or acts of an explicitly sexual nature. It is also possible to take images of clothed bodies and “undress” them, reconstructing what the body would look like under clothing and creating highly realistic images.

Initially, the phenomenon mainly involved famous people for the purpose of discrediting or blackmailing them. But in recent times, with the increasing spread of software using this technology, the risk also involves “ordinary” people, who can become the object of psychologically and socially very damaging actions. Deep nude videos can also be used, without the complete knowledge of the subjects depicted in the images, to fuel the practice of illegal pornography and, unfortunately, even very serious crimes such as child pornography.

 

Right to be forgotten

The right to be forgotten is a key provision of the GDPR: this right aims to ensure that data subjects can have their personal data deleted under certain circumstances, such as when such data is no longer needed for the purposes for which it was collected or when the data subject withdraws their consent for it to be processed.

However, the right to be forgotten may encounter difficulties in practical application in certain cases. Since images and videos created through deep fakes can be shared on multiple online platforms and quickly go viral, the data subject who wishes to enforce this right (and the data controller who has to follow up on the relevant request) may not see their claim fully satisfied. Nevertheless, there are also great difficulties in identifying the origin of the manipulated image. This can make it difficult to determine who created the image and with what intent, making it difficult to identify those responsible and take appropriate legal action.

These issues are just some of the significant challenges arising from the spread and exploitation of deep fakes and deep nudes.

It’s important, therefore, that data controllers, data protection authorities, and companies implementing AI technologies work together to develop effective solutions to address the risks associated with the use of deep fakes and deep nudes to ensure the protection of data subjects’ personal data and guarantee the exercise of the rights recognized by the GDPR. Only through concerted and coordinated action will it be possible to successfully address these challenges and protect the rights of individuals in the digital society in which we live.