Deepfakes: Gender-Based Violence in the Era of Artificial Intelligence

Imagine you are a 16-year-old student living in Latin America. You have an argument with a group of friends or end a relationship, as often happens in adolescence. When you arrive at school the next day, your entire life has changed: you have become a victim of deepfakes containing intimate content as a form of revenge.

With the rise of new applications, this has been a real scenario for many young women in Peru, Ecuador, New Jersey, Brazil, and other countries.

What are deepfakes?

They are the result of using sophisticated technology based on Artificial Intelligence (AI) that allows for the overlay of faces and voices in videos or audio. This creates fake but highly realistic content. In some cases, this tool has been used for extortion, illegal sale of explicit content, manipulation of political speeches, or as a form of mockery towards celebrities and private individuals.

In the early days of this technology, around 2017, the first affected were famous women. Cases like the manipulations of Scarlett Johansson, Gal Gadot, and Jennifer Lawrence raised concerns about the possibility of similar videos becoming a new means to defame and humiliate women.

Subsequently, some websites like Reddit modified their terms of use to prohibit non-consensual explicit content on their platform. However, various websites and forums dedicated to creating and sharing such deepfakes continue to proliferate in cyberspace.

Are deepfakes a new form of gender-based violence?

The growing digital era has increased productivity and efficiency, improved communication, and fostered innovation. However, it has also brought about the emergence of new and sophisticated forms of reproducing other types of violence. Historically, this violence has not affected all individuals in the same way.

An international survey on online harm conducted by the Centre for International Governance Innovation (CIGI) in 2020 in 18 countries with 18,149 participants revealed that 60% of respondents had experienced at least one of 13 forms of online harm. LGBTQ+ individuals reported the highest proportion of experienced incidents. Additionally, women represented a slightly higher proportion than men, though they were more likely to report a more severe impact of these incidents compared to men. Among the online harms surveyed, the non-consensual use of intimate images was the incident most considered harmful, especially for women (82.8% vs. 71.2% for men).

Regarding deepfakes, evidence shows that women are the primary victims. A study on the state of deepfakes conducted in 2019 by the cybersecurity company Deeptrace found that 96% of this type of online videos were of intimate or sexual nature. The victims were primarily women, often actresses, musicians, and to a lesser extent, media professionals. In contrast, videos without explicit content primarily targeted men (61%), mainly politicians and corporate figures.

Gender-based violence in the digital realm

One of the most disturbing aspects of deepfakes is their connection to revenge through the dissemination of intimate and sexually explicit content. This occurs when, without a person’s consent, their partner, ex-partner, or third parties distribute or threaten to distribute sexually explicit intimate images of them with the purpose of controlling, punishing, and/or damaging their reputation. Some research has documented gender patterns in this phenomenon, showing that it disproportionately affects women compared to men, making it another form of gender-based violence.

With increased access and rapid adoption of AI-based technologies, there is a risk that deepfakes become a powerful and dangerous tool for this type of crime, as the perpetrator no longer needs authentic content but can generate it with a quality that increases the likelihood of people believing it is real. On the other hand, victims are left in a vulnerable position, as it is challenging to refute the authenticity of these videos and completely remove them from cyberspace. The repercussions at the social, professional, or health levels can be highly impactful. It becomes evident that the consequences of digital gender-based violence can extend beyond the cyberspace sphere.

Victims of deepfakes may experience generalized anxiety while dealing with shame, anger, humiliation, and stigma. The violation of their privacy and the spread of manipulated images can profoundly affect their mental health and, consequently, their quality of life, and even cause trauma. Additionally, it can have repercussions on the personal relationships and professional opportunities of affected women.

How to protect the safety of women in the era of Artificial Intelligence?

To address this concerning and growing trend, a multidimensional response is required, integrating elements such as legislation, technology, and education.

1.Legislation:

The impact of deepfakes is exacerbated by the lack of adequate legislation and insufficient public awareness and consciousness. In many countries, laws and regulations have not advanced enough or at the same pace as technology. However, there are advancements in related issues, such as Mexico’s Ley Olimpia, which condemns acts of digital violence in multiple forms and has spread to other Latin American countries. Such initiatives make it easier to report this and other forms of digital aggression, enabling effective punishment of those responsible. Promoting information about deepfakes also helps avoid a social environment where these actions are minimized or ignored.

2.Technology

Digital platforms such as social networks and other websites, as well as public security authorities, must implement deepfake detection technologies and other manifestations of digital violence. It is imperative to promote strict policies and actions to prevent the spread of manipulated content and images without the consent of the individuals involved. Moreover, leveraging technology to make relevant reports is crucial. Some examples that put technology at the service of people include:

• LegalTech for LGBTQ+ individuals in El Salvador. These tools facilitate access for people with diverse gender identities experiencing gender-based violence.
• Sistema Salvia in Colombia. This technological tool integrates multiple channels to address gender-based violence: in-person assistance, a helpline, a mobile app, a website, and instant messaging. In this way, various information systems are intertwined, avoiding the re-victimizing process of reporting incidents repeatedly to different institutions.

3.Education

Educational programs must be created to recognize and report manifestations of digital gender-based violence. The educational system, both public and private, can play a fundamental role in this matter. Significant work relates to education in new masculinities, as shown by the Hablemos entre Patas project.

Solutions matching the new technologies

Deepfakes and other forms of online violence represent a growing threat to the security and dignity of women in the current digital world. To prevent the wave from increasing, it is urgent to address this issue comprehensively. Collaboration and coordination among the technological sector, legal and security institutions, and the educational system are crucial to protect the rights and safety of women, among other affected population groups.

The digital world should be a space for growth and creativity where women can express themselves with freedom and security without fear of becoming the next victim of a deepfake.