By Mandar Patil, EVP at Cyble
Visualize a scenario where you receive a call from someone in your family asking for financial help. Their voice appears to be theirs, and they are in panic. You do not get enough time to consider; therefore, you do the transfer. Ultimately, you discover that that person never made that call. This is the new normal era of scams like AI-based voice scams.
Voice-cloning technology enables scammers to easily replicate someone’s voice using available audios. In India, online security authorities are raising the alarm about the increase in reliance on AI techniques in scams where fraudsters impersonate family members, public servants, or other trusted people. Probably the biggest challenge is that one might not distinguish the AI voice just by hearing it. A recent study concludes that people have a hard time differentiating between genuine and synthetic voices according to their findings. It becomes apparent that the old clues like unnatural pauses and other signs signaling robotic sound no longer work.
Then, how to recognize it?
1. Be wary of sudden need
Most times, the first sign is not in the voice but in the situation. Scammers create urgency declaring that important people in your life are involved in various scandals when they were arrested or were at the hospital, etc. The goal is simple, make you act. If you are requested to pay something urgently, pause for a second. In real-life situations emergencies can be verified even for a couple of minutes.
2. Dismiss the voice by itself
Simply hearing a familiar voice is not enough to prove identity. Instead of asking yourself, “Is this voice familiar to me?” you should be asking “Is there any other way I can confirm that this is really the person I think it is?” In case of suspicious calls, it is better to end the call and return to calling the person using a verified phone number you already have in your phone book instead of the number that the call provided.An extra precaution when dealing with the family could be to agree on a question beforehand to confirm the voice.
3. Watch out for suspicious requests
An AI-enabled voice scam may involve requests that appear illogical. For example, you may be asked to perform any unexpected transactions such as moving money to a new bank account or releasing your personal data. Even if a voice sounds like a family member, yet the request strikes you as unusual, you should raise your suspicions.
4. Verifications via other channels
If you are getting involved in a business transaction or making a financial transaction, it is advisable to follow certain verification protocols. Try calling or messaging on their official email address.
It is especially important in business not to trust AI-generated voice when it comes to making payments, changing passwords, transferring money to vendors, etc. No matter how convincing the voice was, it should never override the existing protocols.
5. Don’t fall for AI voice characteristics
Many people assume that AI-generated voices must sound robotic, have strange pauses, and lack any emotions. This is not true anymore as modern phonetics can imitate tone, accent, and emotions in good quality. The best advice here is based on the behavior rather than the technology: pause, verify, and
6. Safeguard the data that scammers may duplicate
Keep in mind how much voice and personal information is easily obtainable on the internet. Videos, interviews, social media posts, and audio recordings can mean opportunities for fraud. It is necessary to always check the privacy settings on social media. Personal data must not be shared for no reason, as it can be used by conmen to make their tales more convincing. Artificial Intelligence does not change the psychological aspects of scams. It has made identity theft much more efficient. Probably, the technology will go further, but the key point to remember is keeping the verification process, and not letting yourself get hurried by urgency, friendship, or fear.