The phone rings. It's your kid's voice. They're crying, scared. They've been in an accident — or arrested — or something's very wrong. They need money wired immediately. They beg you not to hang up.

That's not your kid.

This week, a startup called Savi Security launched specifically to protect people from exactly this kind of attack — fake kidnapping calls and AI-generated ransom demands that clone the voice of someone you love. At the same time, a lawsuit made national headlines claiming ChatGPT had been used to manipulate a vulnerable person into a dangerous mental spiral. And cybersecurity researchers documented what they're calling the first AI-assisted ransomware attack — where artificial intelligence handled the targeting, message writing, and initial contact, automating work that used to require a skilled criminal with hours to spare.

This is not science fiction. This is happening right now, in July 2026. And if you're a veteran, a contractor, or a small business owner who's been skeptical of AI — your skepticism might be the very thing that makes you a target.

The Train Didn't Leave You Behind — The Scammers Just Got There First

Here's what nobody tells you: you don't have to be using AI for AI to affect your life. Scammers are already using it against people who've never touched ChatGPT or Claude. The question isn't whether AI is relevant to you. It already is. The question is whether you know enough to recognize what's happening when it's pointed at you.

Voice cloning is the one that should get your attention first. AI tools can now generate a convincing replica of someone's voice using as little as a few seconds of audio. Where do people keep a few seconds of their voice? On Facebook. Instagram. TikTok. YouTube. A birthday video your kid posted last month. A voicemail you once recorded and shared. A podcast interview from a couple years ago.

Criminals pull that audio, run it through a voice cloning tool, and suddenly they have a digital version of your son or daughter's voice that can say whatever they type into a script. Then they call you at midnight.

Why Veterans and Small Business Owners Are Especially Targeted

If you served, you were trained to act decisively under pressure. You were trained to respond to urgency, to cut through confusion and move. Scammers study behavior profiles, and someone who spent years making quick decisions in high-stakes environments will have a harder time pausing to question a panicked voice that sounds like their kid.

Small business owners face a different version of the same problem. You have more financial accounts, more vendor relationships, and more exposure than the average person. You make money decisions constantly. You get calls from people you've never met in person. When someone calls claiming to be a supplier asking for a wire transfer to a new account, you're already primed to respond — because that's what running a business looks like.

And if you're a veteran who's also running a business? You're doubly exposed. You have the decisiveness that scammers exploit, and the financial access they want to reach.

This isn't meant to scare you off. It's meant to give you the same briefing you'd want before any operation. Here's the threat picture.

The Four Scams You'll Actually Encounter

The fake emergency call. You get a call from a panicked voice that sounds exactly like a family member. They're in danger, in jail, or badly hurt. They need money wired now. Don't call anyone else — they'll say it'll make things worse. Don't call their phone — they'll say it's dead or confiscated. This scam works because your emotional brain overrides your rational brain the moment you hear that voice. The voice feels real because it is — it's a clone built from audio the scammer pulled off their social media.

The vendor or client impersonation. Your "bookkeeper," "insurance rep," or "longtime vendor" emails or calls to update their payment information. You've worked with them for years. Except the email came from a domain that's one letter off from the real one, and the voice on the phone was generated using AI trained on the real person's LinkedIn video or company podcast. Business email compromise supercharged by AI is one of the leading causes of wire fraud losses for small businesses right now. These attacks are personalized, polished, and terrifyingly convincing.

The too-good offer. A "government program" you've never heard of reaches out because of your veteran status. Or a new client shows intense interest in your services and wants to pay a deposit immediately via wire. AI-generated personalized messages can reference real details about your business, your location, your service branch, and your background — making these feel eerily tailored to you specifically, because they were. AI scraped your website, your LinkedIn, and your Facebook and built a profile before the first message was sent.

The AI-assisted ransomware attack. TechCrunch reported this week on what researchers are calling the first AI-assisted ransomware attack on a business. The AI handled target selection, email generation, and initial contact — automating what used to require a skilled attacker with significant time and expertise. Humans were still involved in executing the final steps, but AI is now doing the reconnaissance and social engineering at scale. If your business runs on email and has any web presence, you are a potential target.

What You Can Do About It — Starting Today

The good news is that most of these attacks share a single critical weakness: they require you to act immediately. Urgency is the weapon. Deliberate slowness is the defense.

Set up a family safe word. This is the single highest-impact action you can take in the next five minutes. Gather your family — or text them right now — and establish a word or short phrase that anyone can use to verify they're really who they claim to be during an emergency call. It should be something you'd never say naturally in a crisis, easy to remember, and known only to your immediate household. "The fishing trip was in March." "Grandma's dog is named Biscuit." Something personal and specific. If the voice on the phone can't give you the safe word, hang up and call back on a number you've had for months — not a number they give you on the call.

For your business: build a two-channel rule. Any request to change a payment method, bank account number, or initiate a wire transfer must be confirmed through a second, independent channel. If the request comes by email, call the person directly using a number already in your contacts — not a number in the email. If the request comes by phone, follow up by email to an address you've used before. This one rule stops most AI-enhanced business wire fraud cold. Write it down, post it near wherever your team handles payments, and enforce it without exceptions.

Slow your family's social media voice footprint. You don't have to disappear from the internet. But you should know that every video of your kids talking, every voice memo shared publicly, every Facebook Live you do — all of it is potential source material for a cloning attack. Tighten your privacy settings on family posts. Keep your kids' voices off public platforms where you reasonably can.

Practice pausing before acting. This one is hard if you're wired for quick decision-making, but it's essential. When a call, text, or email creates sudden urgency around money, make it a rule to pause for two full minutes before doing anything. Walk away from the phone. Breathe. Think about whether this is the kind of communication you'd normally receive from this person. Real emergencies are still real in two minutes. Scams depend on you not taking those two minutes.

Using AI to Fight Back

Here's the part that may surprise you: AI can help you protect yourself and your team from these attacks. Not by being smarter than the scammers, but by helping you build the training materials, scripts, and protocols that make your people harder to fool.

Try this prompt:

"I run a small business with [number] employees. Write me a 5-minute verbal training I can walk my team through to help them recognize AI-generated scam calls and emails. Include real examples of what these look and sound like, and a simple decision checklist for what to do when something feels off."

Paste that into Claude or ChatGPT right now and you'll have something you can actually use at your next team meeting. It costs nothing. It takes five minutes to read through, and it might save your business thousands of dollars.

You can do the same for your family:

"My family isn't very tech-savvy and I want to explain how AI voice cloning scams work without causing panic. Write me a plain-English explanation I can share in a family group chat, followed by 3 specific things we should do right now to protect ourselves — including how to set up a safe word system."

Copy that. Paste it. Share what comes out in your family group chat tonight.

The Bigger Picture

It would be easy to read this and decide that AI is just a threat — that your skepticism was right all along and none of this technology is worth touching. That's not the lesson here.

The same technology that lets scammers clone voices is the same technology that helps veterans navigate VA claims without needing to hire someone to decode the paperwork. It's the same technology that lets contractors write professional proposals in a fraction of the time. It's the same technology that gives small business owners a full-time research assistant they can actually afford.

AI is a tool, and like every other powerful tool — a crane, a truck, a rifle — it's being used by both responsible people and criminals. The people most at risk right now aren't the ones using it. They're the people who've decided AI isn't for them, so they haven't learned the basics of how it works. Not knowing how voice cloning works doesn't protect you from a voice cloning attack. Knowing does.

You've been briefed. Now you know the threat. That puts you ahead of most people.

Your One Action for Today

Text your immediate family right now with a safe word. Something specific, something odd, something only your people would know. Tell them: if anyone ever calls in a panic — and they can't say the safe word — hang up and call back directly.

Three minutes. No cost. And it's the kind of low-tech, human defense that no AI on earth can crack.