Il consenso nell’era dell’intelligenza artificiale

For the first time, technology makes it possible to build a credible representation of our identity without our presence, our will, or our participation ever being involved. But if fiction becomes indistinguishable from reality, what remains of our right to decide who we are in the eyes of others?

Deepfakes generated or manipulated through artificial intelligence have produced an unprecedented rupture, separating the physical person from their digital depiction. In doing so, they have added a further layer of complexity to the question of consent, calling the very principle of self-determination into question and further reducing the individual's ability to control how their image is recreated and circulated online.

These audiovisual contents are no longer a mere technological curiosity relegated to science fiction. The distance between what once seemed impossible and what can now be produced artificially has completely dissolved. Today, their most widespread application is non-consensual pornography, objectively a form of digital violence that overwhelmingly targets women. 

According to some recent estimates, including a 2023 report published by

Home Security Heroes, around 98% of the deepfake videos available online are pornographic, and nearly all of the people depicted (99%) are women.

The figure does not merely describe the spread of a technology; it reveals a disquieting trend: the digital appropriation of the female image, turned into an object to be manipulated outside any form of consent.

To understand why deepfakes put into crisis not only technology but also the way we conceive of autonomy and equality, we must start from a seemingly simple question: what protections exist for the right to self-determination?

The right to self-determination is often interpreted as the mechanism by which an individual authorizes or refuses a given act affecting their legal sphere, sometimes ending up reduced to a binary logic: a simple "yes" or "no." Beyond the debates that have challenged this view of consent as a purely dichotomous choice, the conception still rests on the idea of a free and informed decision about a concrete event: something that really happens, an action one can either agree to or oppose.

This is precisely the premise that pornographic deepfakes radically undermine. Unlike traditional violations of the right to one's image or privacy, there is not necessarily a real act in which the person could have taken part, or to which they could have denied their consent. There is no lived experience at all: what exists is a representation entirely fabricated, built independently of the individual's will, yet capable of producing drastic and deeply tangible consequences.

This is where the question goes beyond consent in the strict sense. These images, sexual in nature or not, are no longer about something "done" to a person, but something forced upon them.

When a technology makes it possible to produce a credible version of a person, at times imperceptible from what others perceive as true, the fundamental bond between individual and identity is called into question.

The dimension of sexualization makes the phenomenon even more complex. The problem is not only that the content is false, but that a fiction can nonetheless have concrete repercussions on the person it is attached to. Even when its artificial nature is known, the image continues to operate in social space, shaping perceptions and judgments that depend not solely on its authenticity but on the cultural meaning it carries.

In the Italian context, where narratives of femininity still oscillate between respectability and eroticisation, deepfakes can reactivate stigmas and judgments that already exist. They do not, therefore, create new forms of control over the female body, but amplify pre-existing power dynamics, transferring and adapting them to digital space. 

The case of Giorgia Meloni is an emblematic example, not so much because of the political figure involved as because of the message it conveys.

In 2024, the image of the Prime Minister was used to create and spread online AI-generated pornographic videos.

Beyond its judicial implications, the case shows how the commodification of the body can become a weapon of attack even against the woman who holds the highest office of the State. If not even the highest level of institutional power offers protection from this kind of violence, then the issue goes beyond the single episode. It is a reminder, in some ways a bitter one, that technological progress does not necessarily coincide with cultural progress: despite the strides made toward equality, female identity remains one of the main terrains on which control and delegitimization are exercised, real or artificial as they may be.

Technology introduces new tools, but the logics that guide their use belong to a far older imaginary.

How, then, do we respond to this new form of violence?

As often happens with major technological innovations, the law has arrived late. Artificial intelligence has evolved at a speed that has left wide grey areas in the regulatory landscape, making the need for a new legal framework plain.

The European Union has taken an important step in this direction with the AI Act,

introducing a risk-based system of regulation and new transparency obligations for certain AI-generated content. Alongside it stands the recent directive on combating violence against women and domestic violence, which recognizes for the first time at the European level certain forms of digital violence, including the non-consensual dissemination of intimate material.

Yet, as this phenomenon shows, deepfakes cannot be considered merely a technological problem. 

Legal change is necessary, but on its own it is not enough. The challenge is also cultural: it concerns the value we place on self-determination, on human dignity, and on each person's freedom to maintain control over their own image and their own presence in digital space.