It is said that for the first time in the world, there is a fraud damage caused by synthetic voices made by artificial intelligence. The amount of damage to companies that suffered from fraud amounts to 250 million won.
According to Yuler Hermis, the insurance company of the affected company, the manager of the company received a phone call from his boss in the afternoon. The voice of the boss told me to send the money to a Hungarian bank account because I had to pay extra for the delay in payment. Also, on the phone, he instructed me to send the statement by e-mail. This is actually a synthesized voice that imitates the voice of a boss, but the synthesized voice is said to have not only the same tone, but also tone, word classification, and even German dialect.
After the first remittance, the voice of the same boss called again. The manager was suspicious of this and called the boss again. According to Yuler Hermis, the manager said that the composite voice continued to talk to him even during conversations with his real boss.
In recent years, AI technology that synthesizes video or audio just like a real person has been highly developed. In 2017, it became a hot topic to combine the face of a famous actress with an adult video. In September, in China, ZAO, an app that easily creates deep fake videos with a single face photo, appeared and gained popularity. Of course, concerns are also being pointed out.
In addition, some point out about the danger that false news images can be created indefinitely, spreading fake news and making it difficult to distinguish between truth and lies. Another problem is that if photos or videos spread on the Internet, it becomes difficult to prevent the spread itself even if it is technically possible to detect fake news. Related information can be found here .