DF shares deepfake video of the prime minister: Trend could be 'undermining for democracy', says expert.

Mette Frederiksen appears on TV screens, welcoming guests to a press conference at the Prime Minister's Office.

- After our successful abolition of major holidays, the government has decided to take the next step. We are abolishing Easter. We are abolishing Christmas, it sounds.

The voice is recognizable, but slightly robotic, and a close look at the images reveals that Prime Minister Frederiksen most resembles a video game version of herself.

And rest assured, the government has no plans to abolish more holidays. The video and message are false.

Here is a picture from the video showing a fake Mette Frederiksen. The image is AI-generated. (Photo: © Screenshot from X.)

It is Morten Messerschmidt, chairman of the Danish People's Party, who is behind the video, which was posted on social media X (https://twitter.com/MrMesserschmidt/status/1783882247323492725).

In the video, he dreams of this - false - press conference.

A label in the upper right corner of the video indicates that the content is AI-generated every time the fake Mette Frederiksen appears.

The Danish People's Party has indeed created a deepfake version of Mette Frederiksen, which sounds very authentic and even resembles her, but is ultimately just a computer program.

But if you're a politician, you should keep your hands off creating such AI-generated deepfakes, which resemble and sound like others.

That's what Christiane Vejlø, a digital trend analyst, director of Elektronista Media, and author of the book 'Arguments for People,' thinks.

- Deepfakes are generally very damaging to democracy when politicians are the victims, she says.

Christiane Vejlø has written the book 'Arguments for People,' which puts technological development in a new perspective. (Photo: © Tobias Kobborg, Ritzau Scanpix)

According to her, voters can be misled and might end up voting based on false information.

- In the end, it's undercutting democracy, she says.

Lector in media studies at Aarhus University Jesper Tække also believes that politicians should stay away from such videos and is cautious about their use of deepfakes.

- Our politicians should be moral role models.

- It undermines our trust in them when they use deepfakes in this way, he says.

Read also: Soon, you'll be able to see a new label on Facebook and Instagram. But without pressure, it might not have come, an expert says (https://www.dr.dk/nyheder/viden/teknologi/snart-kan-du-se-nyt-maerkat-paa-facebook-og-instagram-men-uden-pres-ville)

The post has received over 54,000 views on X in just under two days.

Some users write in the comments that it should be clearer that the video is fake or that the Danish People's Party should have refrained from using AI in this way.

Others find Messerschmidt's video amusing and add various smiley faces.

- You can certainly look at individual cases and say, 'Well, yes, that's harmless, or that's just funny,' says Christiane Vejlø.

--------------------

What is a deepfake?

--------------------

A deepfake can be a manipulated image, a false video, an audio recording, or other digital representations created using artificial intelligence (AI).

It is an advanced form of digital forgery, often involving false audio recordings of politicians and other public figures.

Source: The Great Danish Encyclopedia.

However, she emphasizes that one should consider the larger democratic picture if more and more people use the same strategy.

- It's more about the fact that as a politician, you have an interest in ensuring that there is trust in the population regarding what they can rely on, she says.

Christiane Vejlø also believes that part of the problem lies in the fact that the boundaries between political communication and satire have become more fluid, in line with the growing influence of social media.

- Politicians seek attention through their content on social media, and they end up spending more energy on strategically engaging content rather than pure political messages, she says.

Messerschmidt could not be interviewed, but in a written response, he states that the criticism is misplaced.

Morten Messerschmidt has been chairman of the Danish People's Party since January 2022. (Photo: © Liv Latricia Habel, Ritzau Scanpix)

- One must be completely open to humor and critical thinking if one cannot see that Mette Frederiksen's speech is a dream, I dream.

- I have a hard time understanding the criticism.

-----------------------------------------------------

Deepfakes can push the debate towards traditional media

-----------------------------------------------------

Although the video has a satirical tone and the story about Morten Messerschmidt dreaming and a label indicating that the content is AI-generated, Jesper Tække believes it should be clearer.

He had not even noticed the label when he first saw the video.

- Satire has its limits, and it should not be banned, says Jesper Tække.

But according to him, new norms are being established regarding the use of this technology, and Messerschmidt's use of deepfakes contributes to legitimizing such use of AI technology.

- Therefore, it has an unfortunate effect on our perception of what is acceptable in terms of this technology, he says.

>

> Deepfake is pushing us even further in the direction of not being able to trust anything on Facebook, X, or other social media.

> Jesper Tække, lector in media studies, Aarhus University.

Although there are broader boundaries for satire, this type of content requires a debate on where the limits lie, according to Christiane Vejlø.

- It's clearly obvious that there are some broader boundaries for satire, and we haven't fully discussed that yet, she says.

Jesper Tække is concerned that we will see more of this type of content on social media in the future.

Read also: Will you soon be able to watch a romantic movie produced by AI? That's coming soon (https://www.dr.dk/nyheder/viden/teknologi/vil-du-se-en-romantisk-spillefilm-produceret-af-ai-det-kan-du-snart)

The Prime Minister's Office will not comment on the matter.

Several organizations believe that it should be forbidden to create manipulated sexual material - also known as deepfake porn - without the consent of the person whose images are used. You can read more about that here:

Warning: This article was translated by a Large Language Model, in case of doubt, you can always visit the original source.