How it works, in plain language

From any device, email anything you want a second opinion on. Maybe it made you stop and wonder. Maybe it seemed perfectly normal. Scammers design their messages to look ordinary on purpose, so the ones that don't feel off are often the most important to check.

• Email us an email you received.
• Email us a screenshot of a message (text, social media, chat).
• Attach a photo of a document or QR code.
• Attach a voicemail recording.
• Or just email what is happening in your own words.

We will email you back with clear actions.

A reply email with a clear verdict in plain language: SAFE, SUSPICIOUS, SCAM, or UNKNOWN. It spells out what we noticed and the specific steps to take next, so the answer never arrives as a bare label you have to interpret for yourself. Treat the verdict as a strong, evidence-backed opinion rather than a promise: it shows the reasoning it rests on, so you can weigh that instead of taking our word for it.

Before we judge anything, we do the legwork. We follow where a link really takes you, decode QR codes, read the text out of images and PDFs, transcribe voicemails, and check senders, phone numbers, and web addresses against professional threat data. Then our most capable AI weighs that assembled evidence and reaches a verdict. It looks for the same hidden signs a security expert would, like where a link really goes or the way scammers tend to phrase things. Built by cybersecurity experts, using the latest AI technology. No AI is perfect and ours is no exception, so every verdict shows the evidence behind it, and says UNKNOWN when the evidence will not support a call. The free trial is the fastest way to see how it works.

It can happen. There is no perfect AI, just as there is no perfect human: ours can miss something real, or flag something harmless. We would rather say that plainly here than have you find it out at the worst possible moment. So treat a verdict as a strong, evidence-backed opinion rather than a promise, and keep your own judgement in the loop. Every verdict shows its reasoning and the specific things we found, so you can weigh what it is based on. If one looks wrong to you, tell us. We built it, we ship it, and the answer it gives you is ours.

No. There are three things you need to know how to do to get the most out of the service:

1. Forward an email you received.
2. Take a screenshot and share it to an email.
3. Take a picture and share it to an email.

The portal’s How-To section has step-by-step instructions for each, matched to your phone and email app.

We use it to produce your verdict and we don’t sell it, share it, or use it to advertise to you. Your data may inform improvements to the service.

AI can miss things or make things up, and ours is not exempt from that. What we can do is narrow the gap. Before any verdict, we gather the evidence ourselves: we follow links, decode QR codes, read the text out of images, and check senders and web addresses against threat data. The AI reasons from those verified facts instead of guessing. And when the answer genuinely isn’t clear, we built it to return UNKNOWN rather than guess. Borderline messages come back as SUSPICIOUS or UNKNOWN instead of SCAM or SAFE, with the specific things that didn’t add up. It can still get a call wrong, which is why every verdict shows what it found rather than asking you to take our word for it.

Yes. One subscription covers up to 6 mailboxes. Your loved ones forward suspicious messages from their own email and the verdict comes back to them, not you. Everyone in the group sees the subject line and the verdict, but not the contents of the message. That keeps personal mail private while giving the group enough visibility to spot patterns (the same fake-bank scam hitting three relatives in a week) and step in to help when a loved one needs it, without reading anyone’s inbox.

Googling tells you about scams in general. We tell you about your message. Search engines (Google’s AI included) are built to be helpful agents, so they’ll answer even when they should hold back. We’re built to hold back when the evidence doesn’t support a call:

• We read the exact message you sent: sender details, where its links really go, what’s in an image, what was said in a voicemail, the way scammers tend to phrase things.
• We pull in professional threat and research data ordinary search results don’t use: link and QR-code scans, sender and phone-number risk checks, and a safe look at the page a link opens.
• Our most capable AI reasons from that assembled evidence, not from what it happens to recall.
• When the call on your message genuinely isn’t clear, we say UNKNOWN instead of guessing.

We work for you, not the senders. When the evidence shows fraud, we call it. When the sender turns out to be a real, lawfully-operating business, we say so instead of labeling them a scam, even when they use high-pressure tactics. Legitimacy gets checked against independent records: the company's published business address, whether the message’s domain and links actually belong to the brand they claim, and risk data on any callback phone numbers. A real but pushy sender is not labeled a scam; the verdict flags the specific tactics worth watching, like fake-check designs, tracking-only callback lines, and urgency copy, and you decide whether to engage. That is the line we work to draw: real fraud called real fraud, aggressive marketing called aggressive marketing. It is a judgement call, and where the independent records do not settle it, we say so rather than pick a side.

Yes. Every subscription includes a monthly cap on scam checks that is shared across every address on the plan, and the free trial includes a smaller starter allowance so you can try the service before committing. The monthly cap was designed to be enough for family. The exact monthly cap and trial allowance are shown in the portal. Monthly pools reset on the 1st of each calendar month.