© RDNE Stock project/Pexels
No, it’s not even the police who succeeded in tracing Daniela Kleete, a former member of the Red Army Faction (RAF), a small far-left German terrorist group. She had been wanted for attacks, robberies and attempted murder for more than 25 years. It was a journalist from Bellingcat, Michael Colborne, who used a very innovative approach to track down the 65-year-old fugitive. A judicious use of artificial intelligence will have led him to her.
An unexpected lead
The story begins in 2012, where a listener of the podcast « Legion ”, from German public radio RBB meets a woman supposedly called Monika. This meeting took place during a gathering of Anonymous hackers in the city of Cologne.
The woman, whose face appears on wanted notices, presents herself as a former activist of the Red Army Faction. In the summer of 2023, this same listener called the team responsible for the podcast to tell them that this Monika could actually be Daniela Klette.
This testimony is then relayed to Michael Colborne, image verification expert. He begins his research with a facial comparison, but this leads him to a dead end. Determined, he then turned to the Pimeyes software, a facial recognition software.
«& nbsp;I didn't think I was going to find anything, because the photographs in the wanted poster […] had 30 years and that the members of the RAF had remained under the radar for thirty years “, he confided to the Tageszeitung. Against all odds, Pimeyes' research leads him to a woman strongly resembling Klette, taking capoeira lessons in Berlin, who goes by the name of Claudia Ivone. A modus operandi which would not have pleased the European Parliament.
Confirmation and ;#8217;arrest
Continuing his momentum, Colborne obtains irrefutable proof that this Claudia Ivone is not the one ;#8217;she claims. Present until 2019 in capoeira classes, this one is none other than Klette.
The investigation was then published in the podcast « Legion ”, without however this appearing to be 100% certain. Finally, Klette was caught by the police two months later and arrested at the end of February. However, according to the latter, this arrest was made possible thanks to an anonymous tip dating from November, and not thanks to the podcast.
Nevertheless, the Basse-Police Saxe admits to knowing the podcast all the same, which could also explain the search. The Lower Saxony Criminal Police Office told Libération: “ We trusted this source and the clue was followed. Investigations finally made it possible to locate Daniel Klette in Berlin. For reasons related to the investigation, it is not possible to provide further information .” A small difficulty in recognizing the effectiveness of AI if it is not used by an official authority  ;?
- Daniela Klette, wanted for 25 years, was arrested at the end of February.
- Michael Colborne, an expert in image verification, had managed to trace her back using facial recognition software.
- Nevertheless, the German police deny having used the podcast to find Klette.
📍 To not miss any news from Presse-citron, follow us on Google News and WhatsApp.
[ ]