With the rapid progress in artificial intelligence technologies, the digital world is full of opportunities and innovations that promise to bring a significant transformation to our daily lives. However, a new threat is emerging that these technologies are being exploited in sophisticated scams, which makes it difficult for individuals to distinguish reality from deception. Scams have evolved from primitive methods to technological attacks that exploit the potential of artificial intelligence to inflict significant losses on both individuals and companies.
In the past, scams were based on simple methods, such as traditional phone calls made by people reading written texts to convince victims to disclose personal or financial information. As simple as these methods are, they have been remarkably effective. According to the US Federal Trade Commission, Americans lost more than 8.8 billion dollars in 2022 due to fraud, and phone calls were one of the most used methods.
Currently, with the technological development and the use of artificial intelligence technologies, these operations have become more sophisticated and complex than before, which makes them a way to deceive victims very easily, and in a way that is difficult to detect.
How artificial intelligence technologies are used in fraud
The combination of advanced artificial intelligence technologies represents a serious shift in how scams are carried out. Fraudsters have come to use a combination of these technologies to conduct scams The most notable of these techniques are:
Large language models (LLMs):
Large linguistic models can create natural human texts, they can answer diverse questions, give intelligent and quick responses. Scammers can use them to write persuasive texts and generate appropriate responses based on what the victim says, making it difficult to detect that the speaker is not a real person.
Artificial intelligence voice generation:
Currently, AI-powered tools are available that can generate sounds exactly matching the voice of a particular person based on small samples of his voice. Scammers can exploit these tools to make calls that sound like they are from a family member or friend, making victims feel confident in the speaker, and make quick decisions regarding a payment or providing sensitive information.
Generating videos with artificial intelligence:
AI video generation tools can create real-looking videos featuring virtual characters talking and moving in a natural way. And scammers can use these tools to create video calls with the aim of fraud.
Lip movement synchronization:
Technologies such as those developed by Sync Labs are available that can synchronize artificial sound with lip movement in videos. This means that a fraudster can create fake videos in which the characters look like they are really talking.
Analysis of personal data:
Using techniques such as retrieval-augmented generation – RAG, fraudsters can collect and analyze personal information available online. They can use this information to tailor their attacks to look more realistic, such as providing details about the victim's place of work or recent activities.
What do AI-powered scams look like
Imagine that you receive a phone call from a number known to you, in the voice of a family member and during the call he asks you to urgently help with a money transfer or provide sensitive information such as your bank account number. And the scammer may make a video call in which a familiar face appears talking to you, providing accurate details that will make you believe that the matter is completely real. Such a scenario became possible thanks to the integration of advanced artificial intelligence technologies. In the future, these scenarios could become increasingly common, and more complex.
The need for new security strategies:
With the increasing complexity of AI-powered scams, it has become necessary to develop comprehensive strategies to counter them at the individual and corporate levels, the most important of which are:
1-improvement of laws and legislation:
Strict data protection laws can limit the amount of personal information available online, making it difficult for fraudsters to exploit. On the other hand, it may become necessary to regulate access to advanced artificial intelligence technologies, make them available to certain enterprises, and not available to everyone.
2-development of systems to detect content generated by artificial intelligence:
It is important to develop systems that can analyze sounds while making calls to detect fake voice. In addition, it is necessary to develop advanced algorithms for analyzing videos to detect abnormal patterns and fakes.
3-public awareness:
Governments and institutions should launch awareness campaigns to inform individuals about the dangers of these technologies and ways to protect against them.