Robot Harasses Teenage Girl in Cologne: Who Is Criminally Liable When the Machine Reaches Out?
Update Compliance 9/2026
According to media reports, on the night of October 10 to 11, 2026, a robot approximately 1.20 meters (about 4 feet) tall allegedly touched a 16-year-old girl on her buttocks in downtown Cologne. The incident raises a question that is becoming more important for businesses as humanoid robots and AI spread: Who is criminally responsible when a machine violates a legally protected interest?
The Facts
The teenager was standing outside a restaurant in Cologne when the red-and-gold robot, which resembled Iron Man, approached her from behind and eventually touched her buttocks several times. The robot's owner and his companion allegedly filmed the incident and livestreamed it on the internet. Police found the robot in the trunk of a car and seized it, along with the two men's smartphones. They are being investigated on suspicion of sexually motivated insult. Above all, the criminal investigation department is examining whether the robot was deliberately controlled or acted on its own.
Criminal Law Analysis
Under current German criminal law, only a human being can commit a criminal offense. If a robot is operated remotely or programmed to act in a certain way, it is merely a tool. The person controlling it is responsible. If the owner acted deliberately in this case and steered the robot so that it would get too close to the young woman, he may be criminally liable for sexual harassment (Section 184i of the German Criminal Code, Strafgesetzbuch – StGB) or insult (Section 185 StGB).
However, owners, programmers, or manufacturers of artificial intelligence can be held criminally liable only if they are also responsible for the "act" of the AI or the robot. This may be the case, for example, if they deliberately controlled it, programmed it accordingly beforehand, or implemented the relevant behaviors. Liability for negligence is also possible if the person behind the AI breached the required duty of care, for example by failing to take precautions, leaving security gaps in place, or programming it incorrectly, and could have prevented the AI's unlawful conduct.
If no human being bears criminal responsibility, especially where the AI acts so autonomously that its conduct can no longer be attributed to any natural person, a liability gap arises.
Outlook
Legal scholars and the media have long been debating reform of liability (criminal and otherwise) for increasingly autonomous AI. Proposals range from broader criminal provisions targeting the people involved to possible criminal liability of the AI itself (known as "machine criminal law"). Under current law, however, machines cannot be held criminally liable, and the prevailing view (still) rejects the idea. It remains to be seen whether AI will continue to be exempt from liability. Given how quickly artificial intelligence is developing, though, it is almost inconceivable that current law will not be adapted in some form in the future.
Because the legal situation is currently unclear and future developments are uncertain, businesses are well advised to act now. They should clearly assign internal responsibility for each area of AI use, take measures against AI-related breaches of duty, such as issuing policies and conducting training, and document these measures.
