Deepfake pornography: A digital threat and the need for stronger media literacy

In an era inundated with diverse streams of information—from news articles and memes to video content on social media—the ability to think critically has become each individual’s best defense. At the same time, the European Union has recognized the unique dangers posed by the unauthorized creation and distribution of synthetic sexual material. Under the Digital Services Act (DSA), online platforms are explicitly required to remove or disable access to illegal content, including pornographic videos produced without a victim’s consent. Meanwhile, the proposed EU Artificial Intelligence Act places deepfake generation into a high-risk category, mandating strict transparency and human-rights impact assessments. Additionally, the EU’s Code of Practice on Disinformation encourages platforms to self-regulate by detecting and removing false or synthetic media. Despite these measures, technology often advances faster than regulation, and individuals continue to suffer the consequences of non-consensual pornographic deepfakes.

Deepfake pornography relies on deep-learning algorithms—especially generative adversarial networks (GANs)—trained on hundreds or thousands of images and videos of a target individual. Once the model learns the person’s facial features and expressions, it “maps” that face onto the body of an existing adult performer, producing an alarmingly realistic fake. According to media expert Aleksandar Đokić in his report “Deep Fake Porn: Deep Learning and the Fear of Consequences,” this process can take only a few hours, and is typically coordinated on specialized forums and websites that attract tens of millions of monthly visitors.

Although “revenge porn” initially described videos made to shame or retaliate against former partners, today’s deepfake porn often serves commercial interests, political agendas, or simple sensationalism. Media studies professor Stefan Janjić has noted that recent deepfake attacks frequently target public figures and journalists, aiming to undermine trust in their integrity and credibility. These manipulations occur behind the scenes on private servers, where algorithms can precisely graft a celebrity’s likeness into compromising situations.

The harm caused by deepfake pornography goes far beyond personal embarrassment. Victims often experience intense stress, a sense of powerlessness, and symptoms resembling post-traumatic stress disorder. Đokić’s research highlights that women in the public eye—politicians, influencers, journalists—are especially vulnerable, with their bodies used as tools for intimidation and silencing. Even after takedown requests succeed, cached copies and reposted clips continue to circulate, leaving victims trapped in an unending cycle of stigma.

Legally, many European countries—including Serbia—struggle to keep pace with these technological abuses. While criminal codes may punish the unauthorized publication of intimate material, gathering conclusive evidence of AI-fabricated media is challenging, and court cases can drag on for years. Clearer guidelines for verifying authenticity and greater international cooperation are needed to dismantle the networks that create and distribute deepfake pornography.

Combating deepfake porn must be a two-fold effort:

Technological defenses: Develop and deploy detection tools that flag synthetic content by analyzing inconsistent lighting, unnatural eye movements, or audio–video mismatches. A number of startups are already testing models that can pinpoint “unnatural” segments in videos.

Educational initiatives: Integrate mandatory media-literacy modules into high-school curricula, teaching students how AI tools work, how deepfakes are produced, and how to evaluate the credibility of digital media. As Janjić emphasizes, “Education is the only long-term solution to technological deception.”

Deepfake pornography represents one of the darkest facets of the digital revolution: while artificial intelligence can drive creative innovation, it also enables new forms of violence and violations of human dignity. The erosion of truth—manifested in fabricated videos—undermines public trust in both traditional media and social platforms. To meet this challenge, we must simultaneously strengthen our technical defenses and empower every citizen with robust media literacy. Only then can we build resilience against manipulation and safeguard individual rights and freedoms in the digital age.

Sources: Stefan Janjić, Novi Sad School of Journalism https://novinarska-skola.org.rs/ai-u-sluzbi-osvete-i-diskreditacije-ljudi/

Aleksandar Đokić, Fake news Tragač https://fakenews.rs/wp-content/uploads/Dipfejk-porn-duboko-ucenje-i-strah-od-lekcija-2023.pdf

This text is written under the activities of the MEDEA project – Developing Media Literacy to debunk gender related media manipulation and fake news No. 2024-1-LV01-KA210-ADU-000243248  Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them.