INTERPOL has launched an awareness campaign highlighting the growing risk of crime involving AI-generated or manipulated media. Three short videos, also created using AI, demonstrate how so-called deepfakes can be used to commit fraud.

Generative AI applications can convincingly replicate realistic avatars, faces and voices of real people using just a few sources of information. The underlying large language models are becoming increasingly powerful and require less and less technical expertise or investment to use.

Public awareness of the use of deepfakes in criminal activities remains low, according to INTERPOL. The technology is increasingly being used against private individuals as well, for example in investment fraud, identity impersonation, romance scams, sextortion, deepfake pornography and disinformation.

Fraud attempts may take the form of a video call from someone posing as a colleague requesting a bank transfer, a call from a family member claiming to be in an emergency, or an apparently genuine online relationship. The ultimate aim is to steal sensitive personal data or money; INTERPOL particularly warns against unsolicited contacts, time pressure, emotional manipulation, unusual payment methods and virtual relationships that intensify rapidly.

The campaign is based on the motto “Wait. Validate. Don’t take the bait.” It advises people to remain calm and sceptical, verify information independently, agree on a safe word with family members and enable two-factor authentication. Companies should train their teams to recognise deepfake fraud, vishing and social engineering, introduce multi-step verification procedures for financial transactions and data transfers, and keep software up to date; victims should immediately contact their local law enforcement authorities and, if a payment has been made, their bank.