A Canadian investigative journalist sat down with some old photographs one December day in 2023. Thirty minutes later, he had identified one of Germany’s most wanted fugitives. Daniela Klette, a onetime member of the Red Army Faction, had lived openly in Berlin for years. She took part in a capoeira troupe. She appeared in Facebook photos from local festivals. Yet German police never found her. Not until Michael Colborne ran her image through facial recognition software.
The Power of Public Tools Over Police Efforts
Colborne, who works with the open-source investigation group Bellingcat, acted at the request of a German true-crime podcast. The New York Times detailed how the software quickly matched decades-old wanted posters to images of a woman named Claudia Ivone. That was the alias Klette had used while hiding in plain sight in the Kreuzberg neighborhood. Police arrested her in February 2024. They found cash, gold, weapons and fake documents in her apartment. The discovery embarrassed authorities who had spent decades and considerable resources on the hunt.
But the story didn’t end with the arrest. On May 27, 2026, a German court sentenced the 67-year-old Klette to 13 years in prison. She stood convicted of robbery, kidnapping for ransom, attempted kidnapping, aggravated robbery and gun-law violations. Those crimes helped finance life on the run for her and two fellow former RAF members still believed to be in hiding. Additional attempted-murder charges tied to 1990s attacks remain pending. Other terrorism counts have expired under the statute of limitations. Deutsche Welle reported that Klette told the court, “We can only be truly free when everyone is free.” Supporters in the gallery shouted “Freedom for Daniela.” She spoke of her early activism against the Vietnam War and apologized for the trauma caused to victims.
Colborne later told a German newspaper he would have preferred to catch fugitive neo-Nazis. The Slashdot summary of the case, which drew from recent Guardian commentary, noted the contrast between official failure and one journalist’s quick success. And that contrast raises hard questions.
German police had not used facial recognition technology themselves during the long search. Regulations and privacy rules stood in the way. Yet Colborne, operating with publicly available images and commercial software, succeeded where they did not. The New York Times observed that publicly available digital recognition tools could have located her much sooner. Officials called the arrest a milestone. Journalists at the news conference pushed back. One asked pointedly what their success really was. “Listening to a podcast?”
The Red Army Faction, once known as the Baader-Meinhof Gang, carried out bombings, assassinations and kidnappings in the 1970s and 1980s. The group faded but its last known members stayed on wanted lists for decades. Klette, Burkhard Garweg and Ernst-Volker Staub allegedly committed armed robberies to sustain themselves. DNA traces and other evidence linked them, but the three lived carefully. Until one set of social-media photos gave them away.
This case adds to a growing list of successes for facial recognition outside traditional law enforcement. In 2025, UK police used the same category of tools to catch a fugitive who had evaded capture for 27 years. Similar stories surface regularly. Yet the technology remains controversial. Accuracy problems, especially with certain demographic groups, have led to mistaken arrests elsewhere. Civil-liberties groups warn about mass surveillance and the erosion of anonymity in public spaces.
Germany expanded police authority to use facial recognition in terrorism and serious-crime cases in 2024. The timing followed Klette’s arrest but reflected broader European trends. Interpol and other agencies now explore the technology to match faces in social-media posts with terrorist watch lists. The shift feels inevitable. Still, the Klette episode shows the gap between what authorities can do and what determined outsiders achieve with open data.
Colborne’s method relied on comparing old images against large online repositories. No special access to police databases. No classified software. Just persistence, the right search terms and a tool anyone can try. The result landed a former terrorist in court decades after her last known crimes. It also delivered a public rebuke to institutions that had failed to adapt.
Klette’s sentencing closes one chapter. Her two alleged accomplices remain free. Authorities continue to search. They now operate with updated legal powers and, presumably, better technical capabilities. Whether those changes would have shortened the original manhunt remains unknown. What is clear is that one journalist with facial recognition software succeeded where decades of traditional policing did not.
The episode underscores a larger tension. Law enforcement agencies collect vast amounts of biometric data yet face restrictions on how they deploy it. Private actors and journalists face fewer barriers when working with open-source material. The information exists. The tools exist. The question is who gets to use them, under what rules, and with what oversight. Germany’s experience with Klette suggests the current balance may not serve public safety as well as intended.
Recent coverage highlights the sentencing’s significance. The Deutsche Welle account from late May 2026 captures courtroom drama and the lingering shadow of RAF history. It also reminds readers that Klette never fully admitted her role in the group’s final generation. She said she hoped the state still knows little about it. Her conviction on the later robberies shows the long tail of those choices. And it shows how a simple image search brought the past crashing into the present.
Journalist’s 30-Minute Search Ends 30-Year Manhunt for German Terrorist first appeared on Web and IT News.