Deepfakes aren’t necessarily more damaging than hoaxes

We have been dealing with a consequence of digitization for some time now: deepfakes. One problem with this technical achievement is that the faces of known people are placed on other people’s bodies and in this way, for example, false information can find its way into the world. But are deepfakes really such a big threat to our society?

Researchers at the Massachusetts Institute of Technology (MIT) wanted to take a closer look at this. The main question was whether a Video that has been manipulated by deepfakes can cause more damage than a simple text that statically conveys this information to the reader.

7,600 participants were found for the study, who were divided into three different groups. The first group watched a political campaign or a clip from YouTube about COVID-19, while the second team only received a transcription of the videos. The third group served as a control group.

After the videos or transcriptions, the participants received a questionnaire and were asked to assess how credible they found the statements made. In addition, the participants were asked to what extent the content reflected their own opinion.

When it came to what content people believed most, the videos did better. According to the principle “seeing is believing”, people are more likely to think that something has happened when they have seen it with their own eyes. In terms of the content of the statements made, however, the difference was zero.

Because a video convinced the test subjects no more than a written text, both forms can manipulate people equally or not manipulate them. All in all, it can be said that a deepfake is not necessarily more harmful than text with false reports, but what makes the former a bit more dangerous is the way it is presented. Because videos are more likely to be clicked on by users than a long article.

Via Gizmodo