Some parts translated into Englisch below:
Journalists had already tracked down the ex-terrorist, who had been in hiding for 30 years, in Berlin last year. In the rbb podcast “Most Wanted: Where is RAF terrorist Daniela Klette?”, Khesrau Behroz and Patrick Stegemann report on how they followed up on a listener’s tip and, together with an expert from the Bellingcat research platform, pursued concrete leads in Berlin.
The method: Michael Colborne from Bellingcat uploaded old mugshots of Klette to the site Pimeyes - an AI tool for facial recognition on the internet. Private photos apparently showing Klette appeared in the search results.
Klette had been in hiding for 30 years, but Corborne’s search only lasted 30 minutes, according to his own statement. After their research, the two podcasters even visited the Berlin club where Klette had trained for a long time under the name “Claudia Ivone”, says Khesrau Behroz in an interview with ZDFheute. They also discovered a Facebook account for Klette with the false name. However, the trail was lost in the club - Klette had not appeared for training for years, they said.
But if it was so easy for journalists to track down Daniela Klette in Berlin and even find out her cover name - why did the investigating authorities have no trace of the former RAF terrorist for so long? Couldn’t target investigators have simply fed Pimeyes with the mugshots?
In principle, the use of such tools by the police is subject to strict data protection limits. Photos obtained in the course of investigations may not simply be passed on to third parties by the police at any time. When asked by ZDFheute, the Federal Criminal Police Office (BKA) stated that the use of such facial recognition tools is only legally permissible in the context of a court-ordered public search. In Klette’s case, however, this has been the case for decades.
It is unclear whether the podcasters’ findings ultimately led to Klette’s arrest. According to the police, the decisive tip-off came from the public last November.
An interesting Article, why journalists - not police - found her whereabouts:
https://www.zdf.de/nachrichten/politik/deutschland/daniela-klette-verhaftung-gesichtserkennung-pimeyes-100.html
Some parts translated into Englisch below:
Journalists had already tracked down the ex-terrorist, who had been in hiding for 30 years, in Berlin last year. In the rbb podcast “Most Wanted: Where is RAF terrorist Daniela Klette?”, Khesrau Behroz and Patrick Stegemann report on how they followed up on a listener’s tip and, together with an expert from the Bellingcat research platform, pursued concrete leads in Berlin.
The method: Michael Colborne from Bellingcat uploaded old mugshots of Klette to the site Pimeyes - an AI tool for facial recognition on the internet. Private photos apparently showing Klette appeared in the search results.
Klette had been in hiding for 30 years, but Corborne’s search only lasted 30 minutes, according to his own statement. After their research, the two podcasters even visited the Berlin club where Klette had trained for a long time under the name “Claudia Ivone”, says Khesrau Behroz in an interview with ZDFheute. They also discovered a Facebook account for Klette with the false name. However, the trail was lost in the club - Klette had not appeared for training for years, they said.
But if it was so easy for journalists to track down Daniela Klette in Berlin and even find out her cover name - why did the investigating authorities have no trace of the former RAF terrorist for so long? Couldn’t target investigators have simply fed Pimeyes with the mugshots?
In principle, the use of such tools by the police is subject to strict data protection limits. Photos obtained in the course of investigations may not simply be passed on to third parties by the police at any time. When asked by ZDFheute, the Federal Criminal Police Office (BKA) stated that the use of such facial recognition tools is only legally permissible in the context of a court-ordered public search. In Klette’s case, however, this has been the case for decades.
It is unclear whether the podcasters’ findings ultimately led to Klette’s arrest. According to the police, the decisive tip-off came from the public last November.