Origin story · Part 2 of 2

Was it even possible?

Published 29 July 2026 · by Jayme Johnston, founder of Is This A Scam? (a FortifiedWall service)

In part one I wrote down what an answer would have to be: specific to the exact message, backed by real evidence, honest enough to act on, covering everything a scammer might send, never shaming, and awake at 3am. Then I had to find out whether that was a product or a fool's errand.

Isn't this just asking a chatbot?

No, and the difference is the whole reason this exists.

Paste a suspicious message into a general-purpose chatbot and you get an answer that sounds excellent. Fluent, structured, reassuring.

The problem is the effort required to make a decision. How far does the investigation need to go?

Confidently-worded guessing is worse than no answer, because it teaches you to trust the next confident answer too.

How does it actually know?

It goes and looks first. Before anything is judged, the service follows and scans the links to see where they really end up, and takes a safe look at the page they open. It decodes QR codes. It reads the text out of photos and PDFs, so a screenshot of a text message or a photo of a mailed letter becomes something that can genuinely be examined. It listens to voicemail recordings. It checks senders, phone numbers, and web addresses against professional threat and research data — the kind of sources ordinary search results don't draw on.

Only then is the verdict made, by our most capable AI, weighing that assembled pile of verified facts. It isn't recalling what scams usually look like. It's judging what this submission actually is.

And when the evidence genuinely isn't clear, the answer is UNKNOWN, and it says so rather than picking. A service that refuses to guess is a service you can believe when it doesn't.

What can you send the service?

Anything a scammer might use, because covering only one channel would have failed one of my parents. A text (as a screenshot), an email (forwarded whole, hidden information and all), a voicemail, a document — a letter, bill, invoice, or notice, photographed — or a QR code.

Different evidence gets gathered for each, but the standard is the same every time: look first, decide second, and be able to point at why. The verdict comes back by email in plain language with specific next steps, in your own preferred language.

So — was it possible?

After about ten months: yes.

I was aiming for a decent second opinion. What I ended up with is the thing I'd described at the start without believing it was reachable: an honest sounding board that examines one specific message from as many angles as the evidence allows, explains itself, and is awake at any hour so nobody has to feel like a burden for asking.

Could my parents actually use it?

This matters as much as the answer being right. Because it uses ordinary email, it is simple and already familiar.

There are step-by-step walkthroughs, built for the phone and email app you actually have, printable if you'd rather tape a page to the refrigerator door. On newer iPhones: set up the Action button once, and one press sends whatever is on screen.

Try it free