JAKARTA - The FBI issued a stern warning regarding the increasing kidnapping scams that now take advantage of video deepfakes based on artificial intelligence. The perpetrators take advantage of photos and videos spread across the internet, then process them with a genrative AI to create a seemingly convincing fake hostage-taking scene.

The goal is simple and brutal: forcing the victim's family to pay the ransom as quickly as possible before checking its authenticity.

In its latest warning, the FBI explained that these digital criminals collect materials from social media and other online platforms. Once the target identity and face are accumulated, all are processed through AI tools to produce clips that mimic hostage-taking scenarios, complete with expressions and gestures designed to deceive emotionally.

The fake video was then sent as proof of life or rather, proof of manipulation to pressure the family.

AI's quick progress in making verification a challenge. The FBI notes that people now have little time to judge whether the video is real, especially as perpetrators often use automatic deleted messages. However, the FBI mentions several gaps that sometimes appear, for example the proportion of an odd body or the loss of distinctive physical signs that should exist.

Security authorities emphasize the importance of prevention. Residents are asked to limit the personal information they upload, especially when traveling. Privacy arrangements on family photos and videos are also important to be limited to trusted people.

In a threatening situation, the FBI confirmed the importance of trying to contact individuals who claimed to be kidnapped before following the perpetrator's instructions.

This deepfake phenomenon changes the landscape of digital crime. The technology, which was originally developed for entertainment and creativity, is now being used for high-level manipulation, showing how the AI era demands vigilance that is equivalent to its sophistication.


The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)