Deepfakes: Engineering Manipulative Methods to Harm EU citizens

Deepfakes are on the rise worldwide, propelled by Artificial Intelligence. Fake audio recordings manipulated by Artificial Intelligence as a potent weapon used by cyber criminals, were recently created and posted in social media during Slovakia’s elections with the aim of distorting the public image of politicians and the public narrative.

This negative side and unethical use of deepfakes involve cyber criminals exploiting AI to manipulate and create fake audio recordings based on original sources of data like public speeches, but also use original images and videos for that purpose. In the same way a fake audio recording was generated, the same pattern can be followed to create fake images and videos.

Slovakia proved a testing ground for fraudsters that have access to these technologies that can harm citizens not only psychologically but also financially, when the criminals manage to convince potential victims to provide sensitive information, through the use of deepfakes. Their primary goal is to gain their victim’s trust, leading them to provide their bank account details or engage in monetary transactions on fake platforms or websites (cloning).

In a recent successful case carried out by the Greek police in 2023, a major cybercrime gang was busted for allegedly stealing over 6 million euros from unsuspected citizens, committing in total over 270 criminal cases of identity theft. The criminals used advanced techniques to create cloned websites of legitimate banks and businesses, targeting victims with phishing attacks and obtaining e-banking personal data and codes.

As mentioned in a recent article by WIRED (October 2023), Poland and other EU countries can be next. The circulation of AI powered fake content like audio, video, images, means taking proactive measures to counter OIDT by cross checking fake with original sources and distinguishing fakeness. This detection might be easier for image and video deepfakes, but not for audio ones which remain more difficult to prove if they are fake or not, meaning advancements should be made in audio detection technology.

Social media companies like Meta are taking measures to detect and stop these AI generated deepfakes, with new policies and technological solutions implemented in their systems, although criminals are always a step ahead. Law Enforcement Agencies in Europe face the challenge of having to deal with an enormous amount of information and big data in order to mitigate the Online Identity Theft phenomenon.

As previously stated, in response to OIDT, EITHOS EU Project plans to develop tools for detecting and mitigating fake social media botnets, audio, video, image deepfakes created and manipulated by Artificial Intelligence, also by enhancing the capacity of law enforcement agencies to tackle with OIDT (Online Identity Theft).

The EU Digital Services Act is a crucial legislative step in order to counter the growing use of deepfakes and disinformation. Identifying these manipulative tactics in order to protect EU citizens from consuming fake content (video – audio – images) or being tricked by cyber criminals leading their victims to financial loss or psychological trauma.

As the cases of detected deepfakes are proliferating and cyber criminals continue to organize their activity leveraging Artificial Intelligence to harm and deceive EU citizens, the EITHOS Consortium stands to support the fight against Online Identity Theft and reinforce LEAs and Citizens with a strategic & technological context of deterrence against this growing phenomenon.

Author: Efstathios Kassios (KEMEA – Center for Security Studies)





Subscribe to our Newsletters at

Latest News

Subscribe to EITHOS Newseletters

Be updated with our last articles, events and news