On Thursday, September 24th, 2026, many individuals received a call that left them questioning their knowledge on what a typical scam call sounds like. Typically, these calls include a foreign accent, pressure, and a sense of urgency, leading the recipient to believe something is awry. However, a recent incident involving a tech-savvy AI expert has shown that scam calls can be more sophisticated and harder to detect than ever before.
The first call the expert received was supposedly from the police, claiming that his information had been found during a criminal investigation. The caller provided a crime reference number and did not ask for any personal information. Upon checking the telephone number on Google, the expert found several legitimate search results associating the number with the police department the caller claimed to represent. The conversation ended without any suspicious requests.
Later, another member of the supposed police team called and discussed security precautions and potential compromises of personal information. This conversation was helpful, professional, and reassuring. However, it was the third call that raised concerns. The caller mentioned cryptocurrency and recommended installing a security application. The application itself was genuine, but the caller suggested downloading it from a specific website. Upon investigating the website, the expert found it to be a scam, designed to give criminals access to cryptocurrency.
Fortunately, the expert’s suspicions were raised when he noticed the unusual domain extension of the website. He also used his computer to independently check the website and found that even Google Chrome raised concerns about its legitimacy. While he did not download anything, the expert realized that he had come dangerously close to falling victim to the scam.
What sets this scam apart is not just the fake website, but the steps leading up to it. Traditional scammers have always had a time constraint, as they would only profit if the victim handed over money quickly. With the use of AI, scammers can manage multiple conversations at once, building trust over days, weeks, or even months before making a request. This means that the first call may simply be to establish trust, and the subsequent calls can reinforce the story and provide helpful advice.
Another factor that makes this type of scam more convincing is that criminals can manipulate search results when attempting to verify their legitimacy. They can create fake websites using AI, insert carefully chosen keywords into legitimate websites, and even use voice cloning to impersonate someone known to the victim. This means that recognizing the voice of a colleague or family member can no longer be relied upon for authentication.
The most crucial defense against these scams is to verify individuals outside of the conversation they have initiated. If someone claims to be from a legitimate organization, it is best to end the call and contact the organization independently using a number found through a trusted source. For example, if a bank calls, use the number on your bank card or their official website. If a colleague asks for a money transfer, use a different, established company process to verify the request. It is also recommended to have a “safe word” to use when verifying payments. When recommended to download software, always find the official website or app store yourself rather than using the link provided.
The use of AI in scams not only creates convincing voices and fake websites, but it also gives criminals something they have never had before: patience. The scam does not have to happen on the first call; instead, the first call can build trust for subsequent calls. By the time the real request comes, the victim may no longer view the caller as a stranger. This is what makes the AI-powered long con so dangerous.
Juliet Moran, Technical Director of https://telephonesystems.cloud, warns against the use of AI in scams and encourages individuals to remain cautious and verify individuals outside of the conversation initiated. Distributed by
