A fraudulent case in which 25 million yen is deceived by synthesized speech created by AI occurred

by PublicDomainPictures

We found that the world's first scam damage by synthetic voice made with AI occurred. One company was affected by fraud and the damage amounted to 25 million yen.

Fraudsters Used AI to Mimic CEO's Voice in Unusual Cybercrime Case-WSJ
https://www.wsj.com/articles/fraudsters-use-ai-to-mimic-ceos-voice-in-unusual-cybercrime-case-11567157402

Artificial-intelligence voice is used in a theft-The Washington Post
https://www.washingtonpost.com/technology/2019/09/04/an-artificial-intelligence-first-voice-mimicking-software-reportedly-used-major-theft/?noredirect=on

Thieves Used Audio Deepfake of a CEO to Steal $ 243,000-VICE
https://www.vice.com/en_us/article/d3a7qa/thieves-used-audio-deep-fake-of-a-ceo-to-steal-dollar243000

According to Euler Hermes Credit Insurance Company , the insurer of the damaged company, the company's manager received a phone call from her boss one afternoon. The voice of the boss instructed to “transfer money to a Hungarian bank account to pay an additional fee due to late payment” and “send a statement by phone while calling”. This was a synthesized voice imitating the boss, but the voice of Euler Hermes reports that the synthesized voice was not the same only in voice color, it was reproduced to the tone of the voice, the way of dividing the words, and the German accent .

After completing the first remittance, the manager in charge was suspicious because the voice of the same boss called me again, and he called the boss again. In an email addressed to Euler Hermes, the manager wrote: “While I was talking to real Jonas, Jonas, the synthesized voice, kept talking to me.”

In recent years, AI technology that synthesizes images and sounds that look exactly like real people has been highly developed. In 2017, fake porn that synthesizes the face of a famous actress with pornographic video became a hot topic, and a similar technology called Deep Fake has developed from there. In September 2019, an application `` ZAO '' that makes it easy to create deep fake video from one face photo appeared in China and became popular, but after that many privacy concerns are pointed out It has become.

A free app that makes it easy to create deep fake images with a single face photo, popular in China-gigazine

In addition, it has been pointed out that the risk of fake news being spread and making it impossible to tell the truth from a lie by creating infinite news videos.

A movie about Facebook's Zuckerberg CEO who talked about `` control of data '' made by Deep Fake-GIGAZINE

Because it is possible to monetize by diffusing photos and videos on the Internet, it is considered difficult to stop spreading even if fake news can be detected technically. Zeynep Tufecki, a technology-related sociologist, said Facebook and Google's business model 'make people flexible with advertising and create a flow of personal political, personal, and social information' is a problem Pointed out. He said that unless these companies change their business model, problems with the evolution of deep fake technology will remain with the online platform.