"Deepfake Dilemma: The Rising Threat of Video Manipulation in 2024"

by time news
Deepfakes are used to spread false information or create scams on the internet. (Photo AP/Evan Vucci, Archive)

In 2024, deepfake technology has reached a level of sophistication that makes it increasingly difficult to distinguish reality from fiction in videos of public figures. What started as a technical experiment limited to experts has become an accessible phenomenon for anyone with an internet connection and a few digital tools. Figures such as Donald Trump, Elon Musk, and Taylor Swift top the list of celebrities most affected by this type of impersonation.

A study conducted by Kapwing, a company specializing in AI video creation and editing, reveals which public figures are most frequently used in deepfake videos, highlighting this trend that sometimes aims to generate humorous content, but also poorly inform users or directly impersonate someone to carry out scams.

Deepfakes are manipulated videos that use artificial intelligence to replace the face or voice of one person with that of another, in an extremely realistic way. In its early days, creating one required large amounts of data and computational processing, but recent technological advances have made their creation easier. Nowadays, a convincing video can be made with just 10 seconds of original footage and a few clicks, opening the door to all kinds of malicious uses.

"Deepfake Dilemma: The Rising Threat of Video Manipulation in 2024"
Deepfakes are used to spread false information or create scams on the internet. (AP Photo/Julio Cortez, Archive)

The main problem with deepfakes lies in their ability to effectively spread false information and misinformation. In the political context, for example, they have been used to influence public opinion, while in entertainment and social media, they have generated controversy due to the unauthorized use of celebrities’ images. According to the study, “the majority of deepfakes are created to influence public opinion, enable scams, or generate economic benefits”.

Moreover, the fact that these videos can be made quickly and easily, without requiring large amounts of data or specialized training, means that the technology is within reach of anyone. This increases the risk of abuse, as malicious users can create videos that appear authentic but are actually fabricated to deceive or harm.

Kapwing conducted an analysis of the public figures most affected by this technology in 2024. They examined requests for deepfakes made on one of the most active AI tools on Discord, where they manually counted the number of times the creation of fake videos featuring over 500 influential figures from American culture was requested.

Deepfakes are used to spread false information or create scams on the internet. (Illustrative Image Infobae)
Deepfakes are used to spread false information or create scams on the internet. (Illustrative Image Infobae)

The result revealed a ranking of the most deepfaked celebrities, led by Donald Trump, who accumulated over 12,000 manipulated videos just this year. Trump is followed by equally prominent figures like Elon Musk and Taylor Swift, who have also been the subject of multiple manipulations. Below is the list of the 10 celebrities most affected by deepfakes in 2024:

  • Donald Trump:12,384 videos
  • Elon Musk: 9,544 videos
  • Taylor Swift: 8,202 videos
  • Joe Biden: 7,956 videos
  • Tom Cruise: 2,308 videos
  • Dwayne Johnson: 1,970 videos
  • Cristiano Ronaldo: 1,948 videos
  • Mark Zuckerberg: 1,738 videos
  • Will Smith: 1,609 videos
  • Beyoncé: 1,328 videos
Deepfakes are used to spread false information or create scams on the internet. (REUTERS/Hamad I Mohammed)
Deepfakes are used to spread false information or create scams on the internet. (REUTERS/Hamad I Mohammed)

With the growing prevalence of deepfakes, it is important for internet users to know how to detect these manipulated videos. Eric Lu, co-founder of Kapwing, offers three key tips for identifying modern deepfakes:

  • Inconsistent movement of lips and teeth: lip-syncing may fail in details such as the alignment of teeth or the movement of lips. In some cases, the area around the mouth may appear blurry or have unnatural colors.
  • Unusual body movements: deepfakes often focus on the face, meaning that body movements or gestures may not match the dialogue or facial expressions. If the subject does not blink or does so strangely, it may be a sign of manipulation.
  • Inanimated voices and unusual breathing patterns: cloned voices often lack the natural variations of human speech, and pauses or breathing may sound unnatural.

You may also like

Leave a Comment