AI-generated fake videos are circulating on social media, misleadingly presenting news segments from Tagesschau to influence public opinion ahead of the federal elections. One video falsely depicts a swearing-in ceremony rehearsal for AfD candidate Alice Weidel as Chancellor. The clip misuses footage of a military band and employs AI technology to manipulate content. Experts warn that such disinformation exploits the credibility of reputable news outlets, complicating the public’s ability to discern genuine media from deceptive fabrications.
AI-manipulated fake videos are making rounds on social media, falsely portraying news segments from Tagesschau. As the federal elections approach, these deceptive clips aim to sway public opinion.
One particular video claims to feature a tagesschau segment that discusses a supposed swearing-in ceremony rehearsal for AfD candidate Alice Weidel as the new Chancellor. This video is a fabricated piece, generated using artificial intelligence technology.
In the altered video, tagesschau anchor Susanne Holst supposedly announces, ‘The Bundeswehr’s military band is rehearsing the swearing-in ceremony of the new Chancellor, Dr. Alice Weidel.’ The clip includes footage of the military band in front of the Chancellery, performing a rendition of ‘L’amour toujours’ by Gigi D’Agostino.
The song has been co-opted in far-right circles, often associated with racially charged slogans like ‘Germany for the Germans, foreigners out.’ A viral video from May 2024, showcasing young people on Sylt singing these slogans to the same melody, sparked widespread outrage. Even the mere playing of this tune has become a symbol within far-right movements.
The same footage of the Bundeswehr’s military band appears on a YouTube channel, but in that instance, the band performs a traditional military march instead.
This so-called tagesschau video is indeed a fabrication, created with artificial intelligence.
Misleading Claims in AI-Generated Content
The video opens with a genuine segment from ‘tagesschau in simple language. Following the recognizable tagesschau jingle, Holst greets the audience as she normally would.
A reverse image search reveals that this is actually a segment from the tagesschau in simple language aired on June 12, 2024, as Holst wears the same outfit. The forgery only begins from the introduction to the military band segment.
Understanding the tagesschau in Simple Language Broadcast
Upon closer inspection, Holst’s voice has a mechanical quality. Additionally, the term ‘Infomercial’ appears in the upper left corner of the screen. Following the segment, a graphic featuring a quote from Alice Weidel emerges, along with a prompt to vote for the AfD in the upcoming elections.
‘This is an AI-generated video, not a legitimate report from tagesschau,’ stresses Marcus Bornheim, editor-in-chief of ARD-aktuell. ‘The tagesschau is a reputable news outlet and does not engage in advertising.’
The Rise of AI-Generated Misinformation
According to Simone Rafael, spokesperson for the Center for Monitoring, Analysis, and Strategy (CeMAS), this video acts like a ‘self-fulfilling prophecy.’ It presents a desired scenario as though it has already taken place, hoping to influence reality. ‘Poorly informed viewers might start to think that Alice Weidel is already in office. Even those who recognize the video as fake can more readily envision her as Chancellor.’
Reports indicate that the AfD is gearing up for its campaign in collaboration with a far-right agency.
Within AfD supporter circles, deceptive video montages like this have gained traction, often disguised as ‘humor.’ Some are even labeled as satire, but Josef Holnburger, managing director of CeMAS, notes that this label doesn’t always serve as a protective measure against legal consequences. ‘Often, the satire label is a tactic used by disinformation creators to shield themselves under the pretense of artistic expression,’ he explains.
The current video does not carry a satire label. ‘Creating such a video infringes on trademark rights, and we are exploring legal avenues,’ adds Bornheim. The tagesschau brand’s popularity is being exploited with intent. ‘There’s a risk that users may conflate the serious, trustworthy image of tagesschau with election propaganda,’ warns the ARD-aktuell editor-in-chief.
The Credibility Trap of Manipulated Media
At first glance, the altered video appears convincing, remarks Christoph Maerz, head of the ‘Generative AI’ center at the German Research Center for Artificial Intelligence (DFKI). ‘Reputable media outlets like tagesschau hold a unique trust advantage, making them prime targets for manipulation.’
Not only does tagesschau enjoy a high level of trust among the public, but its extensive programming provides ample material for AI training. ‘This allows for the easy creation of convincing AI-generated videos,’ claims Maerz.
The producers of deepfakes are not just using this for electoral promotion. In late 2023, fake audio clips of tagesschau surfaced, misleading listeners into thinking the outlet was issuing apologies for supposed inaccuracies in its reporting.
Currently, fraudulent AI-generated audio files from tagesschau are being disseminated, even at public demonstrations.
Scams Masquerading as Legitimate News
Moreover, fake content purporting to be from tagesschau often circulates, featuring news anchors purportedly endorsing questionable financial schemes. A recent example involved tagesschau presenter Thorsten Schröder, who was falsely depicted promoting a platform by Joko Winterscheidt, claiming that ‘everyone can enhance their financial situation.’
‘Fraudsters behind these fake investment pitches, often referred to as scams, frequently utilize well-known personalities to boost the credibility of their dubious recommendations,’ notes Holnburger. The underlying assumption is that if tagesschau is discussing it, it must be trustworthy.
Old Tactics in a New Format
This strategy isn’t new. Cybercriminals have long employed cloned websites for scams and phishing attacks. Maerz compares these modern deceptions to the classic ‘Nigerian prince’ scam. ‘What once cluttered our email inboxes as spam now appears in video form, enhancing its credibility.’
The sophistication and quality of these fakes are rapidly improving, often promoted as ads on social media platforms. Frequently, they can remain accessible for hours or even days, despite violating the Digital Services Act (DSA) and being subject to rapid removal by platforms across Europe.
This poses a significant challenge, according to Maerz. ‘When deceptive content appears in my social media feed, it becomes increasingly difficult to discern its authenticity and impacts public perception.’