Criminals are using artificial intelligence to simulate kidnappings and extort money from victims through voice cloning and manipulated media [1, 2].

These scams represent a significant escalation in digital fraud because they leverage emotional distress and high-fidelity audio to deceive families into believing a loved one is in danger. By mimicking the exact tone and inflection of a relative, attackers can bypass traditional skepticism and coerce victims into making rapid payments.

The FBI issued a specific alert regarding these virtual kidnappings on Dec. 5, 2025 [3]. According to the agency, fraudsters use AI to create audio of crying or pleading, which makes the simulated abduction appear authentic to the recipient [1, 2]. These tactics are often combined with manipulated photos to further convince the victim that the threat is real [3].

These operations are occurring within the U.S. and globally through various digital platforms [2, 3]. The primary objective of the attackers is to extort money by demanding immediate ransom payments to secure the release of the simulated captive [1, 2].

Law enforcement officials said these scams often peak during specific times of the year. Reports indicate that warnings were also heightened during the 2025 holiday season to protect people from seasonal spikes in fraudulent activity [4]. The use of AI allows criminals to scale these attacks, as they no longer need a physical captive to create a convincing narrative.

Authorities advise the public to remain vigilant and verify the safety of loved ones through secondary communication channels before sending any funds. Because the technology can clone a voice from a short audio clip found online, the risk extends to anyone with a public social media presence [1, 2].

Criminals are using AI to create audio of crying or pleading, which makes the simulated abduction appear authentic.

The shift toward AI-generated extortion marks a transition from traditional social engineering to 'synthetic' fraud. By weaponizing deepfake audio and imagery, criminals can create high-pressure scenarios that trigger a biological fight-or-flight response in victims, reducing their ability to think critically. This trend suggests that traditional verification methods, such as recognizing a family member's voice, are no longer sufficient for security.