German authorities were looking for RAF terrorist Daniela Klette for 30 years. However, AI software for facial recognition provided crucial information after just 30 minutes.
AI facial recognition software provides clues in one of the most famous criminal cases in German history. (Source: ra2studio/depositphotos.com; drizzd/depositphotos.com; image montage: Netzwelt)
- Using facial recognition software, a foreign journalist tracked down a wanted RAF terrorist.
- For the research, he used various online tools that are freely available online.
- It only took him 30 minutes. But such AI programs are controversial.
It took Michael Colborne less than 30 minutes to find a hot lead in one of Germany’s most notorious criminal cases. To do this, the investigative journalist at the international research collective “Bellingcat” used online AI facial recognition software. For 30 years, the German authorities were looking for the RAF terrorist Daniela Klette, who was arrested in her apartment in Berlin-Kreuzberg at the end of February 2024.
However, Colborne had been on the trail of the RAF terrorist months before, as he revealed in an interview with the news website Zeit. The journalist, who specializes in extremism and conspiracy ideologies, researched on behalf of those responsible for the crime documentary podcast Legion: Most Wanted (ARD and rbb). He entered old photos of RAF terrorist Daniela Klette into various facial recognition software.
Colborne leaves it open in the interview which programs are involved, but a look at the transcript of the Legion: Most Wanted podcast provided by rbb reveals that he initially used software from Amazon. This could be the Amazon Rekognition tool. Colborne did not use top secret software for his research, but rather freely accessible databases, photos and AI programs. These may cost a bit of money in some cases, but are potentially accessible to everyone.
Facial recognition software: How it works
The principle is simple: The programs scan entered photos and analyze facial features, special characteristics, head shape, geometric arrangements and even texture properties. Since all of these features differ between people and no two faces are the same, a unique facial profile is created depending on the quality of the photo. The software then searches for this on the Internet, for example in photo databases, on websites or in online forums. In the case of the RAF terrorist, it was successful.
In addition to many mugshots, Colborne quickly found pictures on the Internet that supposedly show Daniela Klette in a sports studio and a German-Brazilian cultural club. Since AI facial recognition does not always deliver error-free results, the journalist checked the images with several programs and also used a facial comparison tool. Ultimately, this check also indicated that the person in the photos found was the wanted RAF terrorist. The research only took 30 minutes.
Apple Photos: How to use facial recognition
All of your images that you have saved on iOS can be scanned and grouped using facial recognition. We’ll show you how you can use and manage facial recognition on iPhone and iPad.
Colborne recommended that those responsible for the Legion: Most Wanted podcast continue to follow the trail of the images found. However, they stopped the research due to a lack of evidence. The State Criminal Police Office is keeping the key whistleblower in the Daniela Klette criminal case secret, but it wasn’t Colborne, as he himself says in the interview. According to him, other journalists were also conducting research at the same time using similar methods and could definitely be considered whistleblowers.
A technology that is accessible to everyone
This case is an example of what is possible with the help of AI and facial recognition in the area of criminal investigation. The fact that such software provides crucial information after just a few minutes about a case that German authorities have been working on for decades is impressive, but also frightening. Because you can see how easily you can track down other people using freely accessible software.
This is also why the use of facial recognition software in Germany is controversial, according to consumer protection. Such programs are viewed critically, particularly with regard to the General Data Protection Regulation. Tools like Amazon Rekognition are officially banned, but DeepFace or Clearview developed by Facebook are not, at least not for private or police purposes.
Theoretically, anyone can use it. The only access requirement is usually that you create an account. With most providers you also have to take out a paid subscription, which often doesn’t cost more than a few euros.
- ” Tip: The best VPN providers for more security and data protection
- » Buy balcony power plant: Comparison of the best solar systems
Don’t miss anything with this NETWORK WORLDNewsletter
Every Friday: The most informative and entertaining summary from the world of technology!