For years, spotting a scam was almost a party trick. Misspelled words. A logo that looked a little off. A story so far-fetched that nobody in their right mind would fall for it. You could glance at an email, roll your eyes, and hit delete.
That era is over.
AI is improving by leaps and bounds, not every year, but practically every week. The same tools we use at DarkHorse IT to write cleaner code and answer questions faster are also available to the people running scams. They use them to write flawless emails, build convincing fake websites, and even copy a voice from a few seconds of audio pulled off social media. The typos are gone. The tells are subtler. And it is getting harder, not easier, to know what is real.
So let’s talk about what still works, and one new trick that levels the playing field.
The Oldest Warning Sign Still Works: Urgency
If you remember one thing from this post, make it this one.
Almost every scam, no matter how polished, leans on the same lever: a sense of urgency. Your account will be closed today. There is a warrant for your arrest. A package can’t be delivered unless you click in the next hour. Your boss needs those gift cards before the meeting ends.
That pressure is not an accident. Scammers want you reacting instead of thinking. The moment you feel your stomach drop and your finger reaching for the link, that is exactly the moment to stop. Real companies, real banks, and real coworkers do not operate on countdown timers. When something demands that you act right now, treat the urgency itself as the red flag.
The Web Address Still Tells on Them
The other classic tell hides in plain sight: the link or the sender’s address.
A message can look exactly like it came from Microsoft, your bank, or Amazon, but the real address underneath rarely lines up. The display name says “Account Security,” and the actual email is something like support@account-verify-now.com. A texted link looks like your bank but points to a domain you have never seen. On a computer, you can hover over a link to see where it really goes before you click. On a phone, press and hold to preview it.
If the name and the address don’t match, that mismatch is your answer.
It’s Not Just Email Anymore
Here is what makes this moment genuinely different. AI scams have jumped off the screen.
We are now seeing cloned voices on phone calls, a grandchild who sounds exactly like the real one, asking for bail money. Fake video. Text messages that carry on a believable back-and-forth conversation instead of a single clumsy ask. The threat is no longer confined to a sketchy email in your spam folder. It can come through the phone in your pocket, in a voice you recognize.
The defense is the same as it has always been, just applied more widely: verify out of band. If you get an urgent call or message, hang up and call the person or company back on a number you already trust, not the one they gave you. A real emergency survives a five-minute pause. A scam usually does not.
Your New Secret Weapon: Ask AI
Now for the part most people don’t know yet.
The same AI that makes scams more convincing is also shockingly good at catching them. If you have a paid account with ChatGPT, Claude, or Google’s Gemini, you can hand it a suspicious message and ask for a second opinion.
Here’s how:
- Take a screenshot of the email, text, or pop-up, or snap a photo of a suspicious piece of physical mail with your phone.
- Open your AI app and look for the little plus (+) button in the chat box. Tap it and add the picture.
- Ask plainly: “Is this a scam or is it legitimate?”
That’s it. In a few seconds, the AI will read the whole thing, point out the warning signs you might have glossed over, and tell you how confident it is. It catches the mismatched addresses, the manufactured urgency, the subtle wording that real companies don’t use. We have tested this on dozens of real-world examples, and it is right far more often than not.
A quick note: this works best on the paid versions of these tools, which support uploading images. The free tiers are improving, so it is worth trying either way, but if you want this as a reliable habit, a paid plan is what makes it dependable.
Think of it as a free fraud analyst sitting in your pocket. When you’re not sure, don’t guess. Ask.
Practice Before It Counts
The best time to learn what a scam looks like is before one lands in your inbox on a busy Tuesday.
That’s why we built a free game: game.darkhorseit.com. It’s called Scam or Legit, and it runs you through three rounds of real-world examples (emails, texts, and notifications) where you have to decide if each one is real or fake.
The catch is the clock. You only get a few seconds per item, on purpose. Real life doesn’t give you all afternoon to study a text message while you’re juggling work, kids, and a phone that won’t stop buzzing. The time pressure is the whole point, because that pressure is exactly what scammers count on. Most people walk away surprised by how a few of them slipped past.
No signup, no account. Just click Start and see how you do. It takes about five minutes, and it’s a great one to run through with your team or your family.
The Bottom Line
AI is going to keep getting better, and so are the scams. But the fundamentals haven’t moved: slow down when something feels urgent, check who a message is really from, verify anything serious through a channel you trust, and lean on AI to double-check the ones you can’t call.
If a message ever makes you uneasy and you’d rather have a human look at it, that’s what we’re here for. Reach out anytime, and catch us every Thursday at 7:40 AM on KFGO 790 AM, where we break this stuff down in plain English.