Deepfake is becoming easier to create thanks to software advancements and the abundance of photos circulating on the internet. According to Reality Defender data from 2023, fake videos made using deepfake technology have surged by 900%. People use deepfake for entertainment, such as pasting their faces onto those of their favorite celebrities or imitating the voices of famous figures. This means that deepfake is no longer considered a difficult technology but rather a new entertaining hobby.
There are at least three types of deepfake that we can find on social media: Face Swap, Voice Cloning Deepfake, and AI-Generated Content. Millions of content have already been created using this technology, ranging from entertainment to controversial materials. Some notable figures who have been deepfaked include Tom Cruise, Elon Musk, Volodymyr Zelenskyy, and Indonesian President Joko Widodo. While deepfake can be fun, unfortunately, it has now become a threat to data security.
In a 2022 survey involving 125 cybersecurity experts, 66% reported experiencing cybercrimes involving deepfake. The UN also issued a report in 2024 showing that cyber fraud in Southeast Asia is becoming more diverse, including online banking fraud, blockchain, cryptocurrency transactions, and other cybercrimes. In Indonesia, a bank reported a case of fraud involving a loan application that turned out to use deepfake.
Currently, deepfake has been misused to deceive the verification process. When registering for an application via a mobile phone, you always go through a selfie verification stage. At this stage, the application system identifies whether you are indeed the owner of the personal data registered. However, this process becomes insecure when your face can be replaced by deepfake. How serious is the threat of deepfake for businesses?
Data from Finextra shows a surge in business losses due to deepfake, from $243,000 to $35 million. However, the impact is not limited to financial losses. Deepfake attacks on businesses can lead to a decrease in consumer trust, expensive reputation recovery costs, and employee layoffs. The average cost of handling data breaches in 2023 increased to $4.45 million, including costs for detection, response, mitigation, and handling the impact of the breach. Unfortunately, many industries lack adequate knowledge about deepfake technology and deception techniques.
So, where is the ethical boundary for using deepfake? So far, the film industry and visual influencers are considered the safest to use deepfake. An example in the film industry is in The Mandalorian (Star Wars series) where the character of young Luke Skywalker was played by Mark Hamill, who was already older at the time.
In the marketing field, deepfake has been used to create virtual influencers. These artificial influencers have appearances and personalities similar to humans, but are entirely created using AI. The cost for virtual influencers is said to be cheaper than human influencers because they are entirely AI-generated. Companies use virtual influencers for marketing campaigns, brand launches, and brand ambassadors.
In facing the challenges and limitations of deepfake technology, it is important for society to remain vigilant and critical of the digital content they consume. While deepfake innovation provides benefits, it is also important to set boundaries to prevent deepfake from becoming a source of crime. Because when deepfake becomes a threat, is it still appropriate to call it entertainment?