Scam alert: Man loses Rs 5 crore through AI-driven fraud

Recently, a man from northern China became the target of a scam employing cutting-edge deepfake technology. As a result of this occurrence, authorities, and the general public are on high alert due to worries that artificial intelligence (AI) techniques could be used to facilitate financial crimes.
 

What persuaded the victim to fall into the trap? 

The said scam took place in Baotou, China. As per the reports,during a video conversation, the scammer successfully impersonated the victim's close buddy using face-swapping technology driven by artificial intelligence. The victim instantly believed the request of his friend who compelled him to send funds for an urgent bidding process, given the resemblance tactic.The naive victim was duped into sending a whopping 4.3 million yuan (more than Rs 5 crore) by the scammer. 

The incident came to light when the victim  talked about the overall situation to his actual friend and when he denied asking for any amount.

The local police said in a statement made on Saturday that they had been successful in recovering most of the stolen money and were vigorously searching for the remaining amount.

China has been actively increasing its monitoring of AI based technologies and applications as a result of its recognition of the growing threat posed by technology driven scams. Moreover, the country, after witnessing a  continuous rise in the manipulation of voice and facial data, implemented new rules in January 2023 to provide legal support to the victims.

What are deepfakes? 

Artificial intelligence (AI) technology known as 'deepfakes' employs deep learning to produce realistic-looking but fake movies and images. 

Large volumes of data are analysed and patterns are discovered by algorithms to produce outputs that are incredibly convincing and realistic. In the case of deepfakes, AI based technology is applied to manipulate videos or images to make them appear as if they involve events or people that never actually occurred or existed. 

The first step in the process is gathering a sizable amount of visual and audio information about the intended subject, typically from publicly accessible sources like social media or official appearances. After that, this information is utilised to train a deep learning model to replicate the deepfake's target.