Deceptive Activities Using Non-I Rico NET in 2025: Methods, How to Identify Them
Sneaky Scammers Levy Unsuspecting Victims Using Fake Neural Networks
Tomsk, May 10th - RIA Tomsk. Research shows that fraudsters increasingly leverage neural networks in their schemes. These advanced technologies allow scammers to falsify voices, create virtual twins, and even deceive the most cautious individuals. Learn how to counteract this progress and avoid becoming a victim in our article from RIA Tomsk.
How do scammers use neural networks?
Voice Manipulation. Fraudsters employ neural networks to copy someone's voice based on a brief recording from social media, voice messages, or phone calls. After acquiring the "sample," they call the victim, pretending to be family, colleagues, or superiors, and demand the transfer of money or access to accounts and credentials.
"Criminals are constantly striving to perfect their deception. Our daily reporting reveals that crimes are being committed using artificial intelligence, including voice alteration, such as modulating tone. For instance, when a grandmother receives a call supposedly from her granddaughter requesting help, this is an old trick that still retains its effectiveness and continues to be used," explained Alexei Vozhov, deputy prosecutor of the Soviet raion of Tomsk.
Chatting with Chat Rooms. "We used to talk about easily deceiving trusting pensioners, but now victims can include people of all ages and social backgrounds, including modern young adults," emphasizes Vozhov.
How can you distinguish a voice generated by AI from a real one?
Unnatural pauses and strange intonation. AI may incorrectly place emphasis or create pauses where a human would never do so. Sometimes, the speech sounds "flat" or too "smooth" without emotion.
Pay attention to pauses within words or between syllables. Also, be wary of monotonous speech or, conversely, an unusually emotional tone.
Uneven modulation of the voice. The generated voice may "float" in timbre or rhythm, especially on long sentences. The start and end of phrases may sound cut off. If the voice seems to jerk, speeds up, slows down, and strange sentence endings or inappropriate intonations appear, this is a reason to be cautious.
Incorrect responses to questions. AI usually responds according to a template and may not understand unexpected questions. A human being may ask follow-up questions or clarify, but AI either ignores or answers strangely.
If you suspect you're talking to a scammer, ask an unexpected personal question. Ask about a pet's name, their class advisor, recall an old memory, or ask them to sing a song or share a personal story.
Poor reaction to interruptions. If you interrupt AI in the middle of a sentence, it may continue uninterrupted, or cut off abruptly. A human being typically responds in a natural way: stops or asks about the issue.
Artificial quality of the sound. Even if the voice sounds realistic, the recording may be slightly "plastic": with inappropriate clarity, the absence of background noise or breathing sounds.
Pay attention to errors in speech: wrong emphasis, strange expressions, missing characteristic phrases, and unusual pauses.
If a video arrives, pay attention to the movement of lips and eyes. The face may slightly lag, the mimicry may be delayed.
Additionally, pay attention to:
- the lack of natural micro-emotions on the face;
- the incongruity between the video and audio, where the voice and lip movements may be slightly out of sync;
What should you do if you're unsure?
Re-contact the person using a known phone number.
Ask surprising questions.
Use video calls: it is harder for scammers to generate videos in real-time.
Delay your decision: if there is pressure to transfer money urgently, this is a reason to be wary.
"Maintaining basic digital and financial hygiene will help you avoid material losses due to scam activities: don't share your bank account details and codes with anyone, be critical of calls from unknown numbers, even those offering easy earnings, and don't install any software at the request of callers, but instead, load anti-virus and anti-spam software on your smartphone for yourself and your contacts," advised Vozhov.
- As scammers advance their tactics, they now employ artificial intelligence, such as neural networks, to manipulate voices and create virtual twins for deceiving victims, even in general news or crime-and-justice contexts.
- In order to detect voice manipulation by AI, one can look for unnatural pauses, strange intonation, poor reaction to interruptions, artificial quality of the sound, and incorrect responses to questions. Additionally, one should pay attention to the lack of natural micro-emotions on the face and the incongruity between the video and audio in a video call.
