Sinister Technology? Scammers use AI-generated human voice to trick friends and family
Scammers use advanced artificial intelligence (AI) technology to create fake conversations with friends and family. This practice is considered complex and sophisticated, representing a new challenge in the fight against fraud and scams.
Photo Reproduction / Internet
Currently, cyber coups have become very frequent. However, with the continuous advancement of technologies, even the most well-known schemes manage to fool many people. A more complex and sophisticated form of coup that has emerged is the use of deepfakes, i.e. scammers are using advanced artificial intelligence (AI) technology to create fake conversations with friends and family.
The coup
With the increasingly easy use of AI programs, criminals have the ability to replicate anyone's voice based on a brief audio sample. Generally, they resort to technology known as deepfake, which allows audio manipulation and simulation of specific voices, including timbre and tone. This makes it difficult for people to distinguish whether they are really communicating with someone they know or not.
The technique, known as "synthesized voice", uses an AI to generate a real-sounding voice that can fool people who know the owner of the voice. These criminals pretend to be friends or family who urgently need money. They ask for the money via bank transfer, hoping that the person will fall for the trap and send money. Victims may also receive fake voicemails from an alleged company employee asking for personal information for an alleged system update.
What is deepfake?
The technology known as deepfake is used to produce fake videos and other types of media, through the use of advanced artificial intelligence (AI) techniques. This technology is capable of creating images so realistic that they portray people performing actions that never occurred or experiencing situations in which they were never present. Through the use of AI algorithms, it is possible to manipulate facial images and create movements, convincingly simulating expressions and speech.
How does the speech synthesizer work?
A speech synthesizer is a device or application which converts text characters into sounds that mimic human speech. The generated sounds can have a somewhat artificial quality, depending on the sophistication of the device used. Speech synthesis can be done by combining segments of pre-recorded speeches stored in a database. However, a new form of scam involves using videos or stories posted on social media as a source for speech synthesis.
For specific applications, the fraudulent individual inserts a script detailing the content and persuasive strategies to be addressed with the victim in an AI tool text box. Then, he issues the command for the artificial intelligence to produce a convincing voiceover, using another person's voice in a realistic way.
Testing the technology
In the last month, the Journalist Joanna Stern of the Wall Street Journal, decided to explore the limits of deceiving people using a virtual clone of their voice and image, created by Artificial Intelligence (AI) systems. The result of the video was not so satisfactory, so she opted to use the cloned voice in a call to her father, requesting a sensitive document, and got an immediate response.
Already the father of Darius Centurion, known as a digital influencer on the Almanaque SOS page, also fell into the trap of his son's cloned voice. However, unlike a simple joke, this call was a scam.
"My father received a call where a person, who he believed to be me, asked for BRL 600. Due to the voice being very similar, my father did not even bother to ask the basics. The person asked for the transfer to an account that was not in my name, but he didn't even care because the voice was mine", said the influencer.
The cloned voice was so similar to the influencer's natural voice that his father, even though he was a musician and sound expert, didn't notice any difference. Centurione, when accessing an application, was able to easily create a replica of his own voice and chose to reproduce the experiment carried out by the Wall Street Journal journalist.
"I went to check if I was fooling anyone and not only did I manage to fool my father again, but also my mother, who is super suspicious. My younger brother realized that the voice was not mine. It is a scam that will become more sophisticated, in time soon the tools should be faster and more efficient, which can make older people victims more often.
What the authorities say
Police warn the public to take note of these phone calls and verify that they are really the people they claim to be. They also advised not to share personal or banking information over the phone or text message. The police emphasize the importance of being vigilant in relation to the veracity of the information. Before taking any action based on a phone call, message or video, it is essential to verify the information in other sources. For example, it is recommended to call back, confirm by email or consult with other people who can guarantee the authenticity of the information.
Law enforcement officials also recommend that you protect your data by reducing the amount of personal information you share online. By providing less data, you limit the material available to cyber criminals looking to create deepfakes or imitate your voice. It's important that youfamiliarize yourself with technology and keep up to date) about the latest fraud trends and tactics to properly protect yourself. Also, when sending money to a family member, always check the recipients carefully, ensuring the security of your bank account.
