ANKARA

Scammers have been defrauding people by using artificial intelligence (AI) video generators to add fake voices and facial expressions to footage of well-known figures such as American billionaire Elon Musk, former US President Donald Trump and Amazon CEO Jeff Bezos.

They try to steal people's bank information and investments, trapping their victims through AI-generated audio and video called deepfakes.

The targets of the scams are often elderly people who are interested in cryptocurrencies but have limited knowledge about safe internet use.

Through these fake videos, where famous names offer "radical investment opportunities promising quick returns" and talk about the economy, people are asked to invest in non-existent companies.

Videos that can be created with a very low budget can be prepared in minutes and published as advertisements on popular platforms such as Facebook and YouTube.

In the videos, where the accents of famous people can be imitated, fake graphics and news site logos can also be used to mislead people.

Musk is the most preferred person in this scam method, with more than 2,000 deepfake videos detected.

According to experts, deepfake videos are not "completely perfect" but are improving and becoming more convincing.

Lou Steinberg, founder and a managing partner at CTM Insights, a research lab and incubator focused on solving structural problems in cybersecurity, told The New York Times that the number of people resorting to this method will increase day by day as organized crime groups realize that they can misuse this technology for their own purposes.

A recent New York Times report cited 82-year-old retiree Steve Beauchamp as one of the victims of deepfake scams. Beauchamp lost most of his investments after being fooled by a deepfake Elon Musk advertisement in 2023.

Also, the science journal Nature shared the stories of South African urologist Kgomotso Mathabe and Indian doctor Viswanathan Mohan.

Mathabe realized that a deepfake video of her was being used in an internet advertisement after phones at her practice were swamped by angry callers accusing her of stealing from them.

She said that viewers of the video were directed to a website, where their account information was stolen.

Mohan also discovered that people were being scammed with a video created on the internet using his face and giving expert opinions in a language he did not know.

He said it poses a legal problem as well as amounting to reputation assassination.

While YouTube said it had removed more than 15.7 million channels and 8.2 million videos for violating its guidelines from January to March this year, Facebook says it has been using a combination of automated and manual systems to help detect scams on its platform.

*Writing by Gozde Bayar