column
Fried – view from Berlin
Scholz or KI-Scholz? When deepfakes become a threat to democracy
What does it mean for democracy when we no longer know what is original and what is fake?
© ZFPS/Youtube
A fake video of the Prime Minister recently caused an uproar in Denmark. How will we find out in the future whether video messages are real? For some people you just hope that they aren’t, says our author.
Recently the Federal Chancellor spoke to us again via video. Olaf Scholz had good news: the Germans are hardworking, and with him as head of government there should be no increase in the retirement age. Sounded really mischievous.
To record the video, the Chancellor casually opened his white shirt collar and wore his usual, somewhat grim look. After about 30 seconds, the camera jumped even closer to the Chancellor’s face. Everything looked real. But was that it?
A similar video with the Prime Minister has now caused a stir in Denmark. However, Mette Frederiksen told her compatriots the opposite of the Chancellor’s message: the Danes were not working enough. In Denmark, all public holidays should therefore be abolished, including Easter, Pentecost and Christmas. Only the Muslim holiday of breaking the fast will be retained.
The German video was real, the Danish one was a fake, made by the head of the right-wing populist party – with the help of artificial intelligence. In this way he protested against the actual abolition of a public holiday in Denmark. The radical message of the video sounded patently absurd. There was also a reference to the AI origins at the top right. Nevertheless, the “Süddeutsche Zeitung” reported, many Danes thought the fake was real. What does this teach us?
There is no easy protection against deepfakes
First of all, the video of the right-wing populist that was circulated on social media was probably seen by a particularly large number of his followers, which gives an impression of their less developed intelligence. Party leader and video distributor Morten Messerschmidt seemed a bit embarrassed. If you don’t recognize that it’s satire, he defended himself, “then it’s not me who has the problem.” That’s correct. However, the problem of some could well become a problem for all of us at some point.
Because what if nothing is displayed in the top right anymore? What if a video message from the Chancellor (for example: “I admit that, for once, I made a mistake”) might seem strange, but still just possible? To put it bluntly, what does it mean for democracy when we no longer know what is original and what is fake?
The European Union wants to use AI law to ensure that so-called deepfakes, including videos made with artificial intelligence, must always be labeled. However, it is difficult for me to imagine that everyone would adhere to this under the protection of internet anonymity. The only real protection remains doubt and the ability to judge a Chancellor’s message based on its plausibility, even if one cannot check its authenticity. In order to maintain this ability, it helps if you regularly inform yourself – small advertising block – in reputable media.
Now it is our great strength star, to see not only the danger in new developments, but also the good. Markus Söder, for example, made a video around Easter in which he offered a large chocolate egg with his likeness for a raffle. In such cases, it’s good that you can now convince yourself that it’s definitely an AI-generated video. This would at least give rise to the possibility that there is some form of intelligence behind it.