‘In five years 90 percent online content will be manipulated by deepfakes’

‘In five years 90 percent online content will be manipulated by deepfakes’

Deepfakes will influence the political and social debate more often in the future. The technology can be used to spread fake news, influence elections and public opinion and even create pornographic images. Experts believe that in five years’ time, more than 90 percent of all content on the Internet will be partially or completely manipulated.

This is apparent from research by Tilburg University of Law, Technology and Society (TILT), which was carried out on behalf of the Scientific Research and Documentation Center (WODC). The findings and predictions about the influence of deepfakes can be found in the report ‘Deepfakes. The legal challenges of a synthetic society.

Deepfakes in a nutshell

A deepfake is a photo or video that is fake but looks lifelike. New digital content is created through a combination of machine learning and artificial intelligence (AI). This consists of existing video images, voices and images. The technology to make deepfakes is getting better every day. This makes it more difficult every day to distinguish the real from the fake.

Deepfake technology has been used in Hollywood for quite some time, for example, to portray a young Luke Skywalker in the Star Wars series The Mandalorian. Thanks to technology, Will Smith battles his younger self in Gemini Man, and you’ll see Robert De Niro and Al Pacino grow older and older in The Irishman. Last Christmas, this video, in which Mark Rutte, Hugo de Jonge, Diederik Gommers and Jaap van Dissel supposedly sing a Christmas song, even went viral.

Negative outgrowths of deepfakes

Deepfakes are already widely used for entertainment purposes. But the technology is not so harmless, scientists warn. For example, deepfakes can also be used to influence public opinion or the outcome of elections. Suppose a video suddenly appears in which outgoing minister De Jonge (Public Health) announces that corona vaccines are bad for health. Citizens can therefore decide to no longer vaccinate against the coronavirus, with all the consequences that entail for healthcare and society.

Deepfake technology can also be used to create or manipulate ‘evidence’. If a judge cannot see through that, it can lead to wrongful convictions. Deepfakes are already widely used to make erotic videos. Presenter Dionne Stax has experienced this firsthand. Although she did not play a part in the porn film, her face was used. It could have ruined her TV career. Spreading such images can have a major impact on a person’s private life and self-image.

As deepfake technology continues to improve and become more widely available, experts fear that in five years more than 90 per cent of all content on the Internet will be partially or completely manipulated. “It’s not just about deep fakes, but also about small manipulations. So at a certain point you no longer know what is real and what is fake, and if something is fake: what exactly? That can cause a lot of problems for journalists or judges,” privacy and big data expert Bart van der Sloot of Tilburg University predicts to BNR.

More regulation

In the report ‘Deepfakes. The legal challenges of a synthetic society, the researchers argue that current legislation and regulations already impose many rules on the use of deepfakes. However, that is not enough to form a fist against misuse of this technology. According to the authors of the research report, more regulation is needed to really do something against the negative excesses of deepfakes.

A ban on the production, offering, use and possession of deepfake technology for the consumer market is an example of this. “After all, current law does not restrict the development of deepfake and its offering, but its use for specific purposes. The damage has already been done by then,” the researchers warn.

More awareness

Van der Sloot believes that politicians should not only focus on regulation, but also on awareness. He acknowledges that this is easier said than done. He says to BNR: “You have to use one technique to scan for digital manipulation, but you also have to make people very aware of the possibilities. Everyone knows the stories of children calling their parents and asking for money. In such a case, a smart parent will switch to video calling, but that too can be manipulated nowadays. At a certain point you can no longer see the wood for the trees, and having journalists check everything completely broken is also impossible.”

The researcher at Tilburg University says that technology currently recognizes 65 per cent of all deepfakes. That will decrease in the future. “You will have to counteract the rest through prohibitions or awareness. That is why it is best to look at how you can prevent such a large flow of deepfakes from arising,” says Van der Sloot.

Catch up on more articles here

Follow us on Twitter here

Popular

Must read

MORE ON THIS TOPIC:

Related Posts