Protocol 1: The Family Safeword
The single most powerful tool against a voice-cloning scam is one that artificial intelligence cannot replicate: shared knowledge. Establish a family “safeword” or “codeword.” This is a unique, memorable word or short phrase that only your immediate family members
know. The rule is simple: if anyone claiming to be a family member calls in distress and asks for money, the first thing you do is ask for the safeword. A scammer, no matter how convincing their AI-generated voice is, will not know the word. They may try to deflect, get angry, or heighten the sense of urgency, but if they cannot provide the word, you hang up immediately. How do you choose a good safeword? Pick something that is not easily guessable from your social media or public records. Avoid pet names, street names, or common words. Instead, think of an inside joke, a memorable but obscure holiday detail, or a pair of random, unrelated words like “orange blanket” or “copper kettle.” Share this word in person and ensure everyone, including elderly parents and adult children, understands its importance. This simple practice acts as an immediate emotional circuit-breaker, replacing panic with a clear, logical verification step.
Protocol 2: The Mandatory Callback
AI voice scams are built on a foundation of urgency. Scammers create a high-pressure scenario—a supposed accident, arrest, or medical emergency—to short-circuit your critical thinking and push you to act before you can verify the story. The mandatory callback protocol is designed to break this cycle. The rule is absolute: no matter how urgent or convincing the plea, you hang up the phone. Then, you immediately call your family member back on the number you already have saved in your phone. Do not use a number provided by the caller or hit redial. This simple action accomplishes two things. First, it moves the conversation to a secure channel that you control. A real emergency will still be an emergency after a 60-second callback; a scam will not survive it. Second, it instantly exposes the scammer’s tactics. Fraudsters often use “spoofing” technology to make it look like the call is coming from a trusted number, but they cannot actually receive calls on that number. By hanging up and initiating a new call yourself, you bypass their entire setup. If you can’t reach your loved one on their primary number, try another family member or a trusted friend to verify their status. This protocol puts control back in your hands, taking away the scammer's most potent weapon: manufactured panic.
Protocol 3: The Un-Googleable Question
While a safeword is the gold standard, it helps to have a backup layer of verification. This is where the personal, un-googleable question comes in. If you are ever in a situation where a safeword wasn’t established or is forgotten in the heat of the moment, you can deploy this technique. The goal is to ask a question that only the real person would know the answer to, and which could not be found through a quick search of their social media profiles or public records. Bad questions include: “What is your dog’s name?” or “What city were you born in?” This information is often publicly available. Better questions are rooted in shared memories and specific, non-public details. For example: “What was the name of that horrible dish I tried to cook for your last birthday?” or “What was the one thing we always argued about on our childhood vacations?” Even a simple, “What did you have for breakfast this morning?” can be effective, as it’s a detail a scammer would have no way of knowing. An AI can clone a voice, but it cannot clone a lifetime of memories. If the caller hesitates, gets the answer wrong, or tries to evade the question, you have your answer. It is not your loved one. This protocol, like the others, relies on creating a small moment of friction that a scammer cannot overcome, giving you the clarity needed to identify the fraud and hang up the phone for good.
