New video series on the "No likes for lies" campaignExpert talk on the topic of "Deepfakes and extremism"

The use of generative AI tools now makes it very easy to create fake images, videos or audio recordings of people - so-called deepfakes. Deepfakes can be particularly dangerous if they are created and distributed by extremist groups for the purpose of targeted disinformation. For example, to create a political mood. In the klicksafe expert talk on "Deepfakes and extremism", we talk about this with Dr. Bernd Zywietz. He is head of the Political Extremism department at jugendschutz.net.

What are deepfakes?

The Pope in a thick white down jacket, Donald Trump forcibly taken away in handcuffs and Olaf Scholz applies for an AfD ban. There are many fake images and videos of well-known people, so-called deepfakes. With some images, it is easier for viewers to recognize that they are fakes. For example, because a person has too many fingers or teeth. With others, however, the fake is almost impossible to recognize. This is because deepfakes, which can be created using generative AI, are getting better and better. Dr. Bernd Zywietz from jugendschutz.net explains what we can understand by deepfakes and what we should know about their origin.

→ The video is also available for free download as an mp4 file.

How can deepfakes be used by extremist groups?

Not every deepfake originates from the targeted dissemination of disinformation. However, such recordings become problematic and a risk for society when extremist groups share such content in order to incite society against individuals and stir up hatred, for example. Dr. Bernd Zywietz deals with political extremism and knows how deepfakes are used in this context.

→ The video is also available for free download as an mp4 file.

What effect can deepfakes have?

Deepfakes can evoke strong emotions, for example stirring up outrage or hatred against individuals or population groups. And especially before elections, deepfakes can be deliberately spread to polarize society, as Dr. Bernd Zywietz from jugendschutz.net has observed.

→ The video is also available for free download as an mp4 file.

How can young people come into contact with extremist deepfakes?

Extremist deepfakes are often first shared in groups, for example on the messaging app Telegram, before they end up on major social media platforms such as Instagram and TikTok. Once spread, extremist content is difficult to recapture. This means that children and young people can also come into contact with deepfakes that pursue extremist goals. In the video, Dr. Bernd Zywietz from jugendschutz.net explains why young people can recognize manipulated content more quickly than older generations, but why there is also a risk that they will not classify the content correctly.

→ The video is also available for free download as an mp4 file.

How can we deal with deepfakes?

In future, it will become increasingly difficult to recognize whether a photo, video or audio recording could be a deepfake. Some services stipulate that AI-generated content must always be marked as such. However, not all deepfakes are actually flagged as such. Dr. Bernd Zywietz gives us important tips on how we can deal with deepfakes on the internet today and in the future.

→ The video is also available for free download as an mp4 file.

Further information on "No likes for lies"

The klicksafe expert talk on "Deepfakes and extremism" is part of the klicksafe campaign "No Likes for Lies" for Safer Internet Day 2025. The aim of the campaign is to sensitize young people to extremism and populism on the Internet. All information about SID 2025 can be found on our SID landing page.

You are also welcome to use the other materials relating to Safer Internet Day: