Italian Prime Minister Georgia Meloni has recently taken a unique step. He has applied to the European Union Intellectual Property Office (EUIPO) to trademark his voice. This has been done so that no fake video or audio of them can be made with the help of Artificial Intelligence (AI).
This news has again brought the growing threat of deepfakes into discussion. Let us understand what deepfake is and why it has become so dangerous.
What is deepfake?
Deepfake is a technology in which with the help of AI, a person’s face, voice or entire video is made so real that the common man considers it to be real.
The word ‘deep’ comes from deep learning, which is a method of AI. In this, the computer learns millions of pictures and sounds and then creates fake content.
For example, by recording a few seconds of a leader’s original voice, AI can convert it into a full speech or song-like sound. Even in a video, someone’s face can be pasted on another person’s body. Nowadays, this technology has become so cheap and easy that by using only 3 seconds of real voice and spending about 30 Euros, such an audio deepfake can be created, which 70 percent people accept as real.
How does deepfake work?
First the AI model is trained. In this, many samples of real voice or face are given. Then AI creates new content by understanding those patterns. Voice cloning occurs in audio deepfake.
Face swapping or lip-sync technique is used in the video. Earlier this required expensive software and experts, but now with free apps and online tools anyone can create deepfakes sitting at home.
Why is Melonie worried?
Georgia Meloni has been a victim of deepfakes before. His photo was doctored to show him in underwear, which caused a huge controversy. The name of his autobiography is ‘Io sono Giorgia’ (I am Georgia). Now she has submitted a 4-second audio clip in which she says twice – ‘I am Giorgia Meloni’. If the trademark is approved, she will get the exclusive rights to her voice throughout Europe. No one else will be able to use his voice without permission. This will protect against fake content and fraud. Trademark is available for 10 years and has to be renewed later.
Threat of deepfakes in the world and India
According to Kaspersky’s report, crimes of voice cloning are increasing rapidly. During the 2024 Lok Sabha elections in India, deepfake videos of Narendra Modi, Rahul Gandhi and Amitabh Bachchan were widely spread on WhatsApp, TikTok and X. The Election Commission had to issue an advisory against this.
Deepfakes are being used to spread rumours, cause political harm, blackmail or commit fraud. Melonie is not alone. In Italy, singer Giusy Ferrari and actor Luca Ward have also trademarked their voices. This shows that celebrities and leaders are now taking legal steps to protect their identity.
What next?
With the increasing reach of AI, it has become easier to create deepfakes, but this is also increasing distrust in the society. People have started wondering whether what they are seeing or hearing is real or not. Steps like trademarks provide protection, but better detection tools, strict laws and awareness are also necessary to completely stop it.
This decision of Georgia Meloni reminds us that in the digital age, protecting one’s voice and face has become as important as protecting a passport or Aadhar card. Deepfakes are not just a source of entertainment but also a serious threat. Therefore, it is everyone’s responsibility to be careful and spread correct information.
