Your spam filter asks who sent it.
We ask what it wants.
The scams that actually take money from people are not badly written, and they do not come from senders anyone has flagged. They arrive from real accounts, pass every security check, and mention your bank, your grandchild, or the invoice you were already expecting. Here is what we do instead, in plain English.
Takes two minutes · Nothing to install · Or start your free month
The dangerous ones pass every check.
Filters were built for junk mail sent to millions of people at once. Today's scam is written for one person, sent from an account that genuinely belongs to someone, and carries nothing a scanner can call malicious — no virus, sometimes not even a link. All the traditional checks come back clean, because there is nothing there for them to catch.
Every automatic check says deliver it. And it is a scam — the domain is not the bank's, the "fraud department" number goes to the scammer, and the whole message exists to get a person to read a code aloud.
Three questions a filter never asks.
Strip away the logo and the wording and almost every scam has the same skeleton underneath. It is asking for something, it got here somehow, and it was written to be believed by this person. We resolve all three for every message.
What is it asking for?
Money moved, a login, a one-time code read aloud, gift cards bought, a new bank account used for a payment — or, the giveaway that outranks all of them, secrecy. Legitimate mail almost never asks you to keep it from your family.
How did it get here?
A domain registered last week. A letter swapped in a familiar name. A reply pasted onto a conversation that never happened. A QR code standing in for a link. A payment page hidden inside a PDF or a picture. We read all of these, including the text inside images.
Why would it work on this person?
It names a real grandchild. It arrives the week a policy renews. It follows a scam they already fell for, which is the single strongest predictor of the next one. A message is only dangerous in the context of the person receiving it, so that is where we judge it.
Five layers. Every message, every time.
Each layer catches what the one before it cannot. Most mail is settled in the first second and never costs anyone a thought; the small remainder that is genuinely ambiguous is where the careful work goes.
Who really sent it
Cryptographic sender checks, the true age and reputation of the domain, whether the sending computer is on a live threat list, and whether the display name is quietly pretending to be someone you know.
What it carries
Every link followed to where it truly leads, including through redirects and shorteners. Attachments opened and read — inside PDFs and Office files. QR codes decoded. Malware and ransomware scanning, plus live threat-intelligence feeds refreshed daily.
The con itself
An AI analyst reads the message the way a person would and judges the intent: what is being asked, what pressure is applied, what story is being told. This is the layer that catches a scam nobody has ever seen, because it never needed to have seen it before.
Against your normal
Who this person actually corresponds with, which senders are established, what a normal week looks like for them. The same message can be routine for one household and alarming in another, and we weight it accordingly.
A human being
Serious detections are reviewed by a working security architect before anyone is alarmed. Software does the reading. A person owns the judgment, and answers when you call.
Quiet, until it matters.
Setup, about 15 minutes
Sign in the same way you already sign in to your email and grant read-only access. Nothing to install, no passwords shared with us, no change to how mail arrives. We do it with you on the phone if you would rather.
An alert that explains itself
A dangerous message triggers a text or email within minutes — in plain English, saying what the message wants and why we think it is a con. It goes to the person and to the family members they chose.No one has to confess
A weekly summary
One short note covering everything reviewed and everything caught, so you can see the watch working without having to think about it in between.
Nothing is deleted
We do not quarantine, hide, or reply to anything. We are read-only by design, so the decision always stays with your family — we make sure it is an informed one.
What we will not do.
- We cannot send, delete, or change anything. Read-only is enforced by Google and Microsoft at the account level, not promised by us. Disconnect in one click and nothing about the mailbox changes.
- We do not ask for your email password. You sign in with your provider; we never see it.
- We do not sell, share, or train on anyone's mail. Not to advertisers, not to data brokers, not to anybody.
- We do not pretend to catch everything. Our Scam Library marks every scam type as fully detected, partly detected, or not yet — and the "not yet" entries stay on the page.
- We do not lock you in. Published prices, first month free, cancel any time, no contract.
Read our methodology → · What we detect · Trust & permissions
Where our watch ends.
We watch email. A scam that happens entirely on the phone, in a text message, or inside a social-media app is outside what we can see — though the email that so often comes before or after it is not. We would rather tell you that here than have you find out at the worst moment. If you want a fuller picture of your exposure before deciding anything, the free check costs nothing and asks for nothing.
See what is already sitting in the inbox.
The free check takes about two minutes, installs nothing, and changes nothing. If you would rather talk to a person first, call or text — a real security architect answers, no obligation, no sales team.
Run the free check →Call or text (940) 281-6672 — a person answers
Common questions
How is this different from the spam filter I already have?
A spam filter mostly asks whether a message looks like junk it has seen before, and whether the sender has a bad reputation. The scams that cost people money are written one at a time, sent from clean accounts, and pass every one of those checks. Family Sentinel reads what the message is asking the person to do — move money, share a code, keep a secret — and weighs that against who the sender really is and how that person normally gets mail.
Can Family Sentinel read all of my email?
We request read-only access, which Google and Microsoft enforce at the account level. It does not include sending, deleting, replying, or changing anything. We look at messages to judge whether they are a threat, and you can disconnect us in one click, at which point nothing about your email changes.
What happens when something dangerous arrives?
A high-risk message triggers an alert within minutes — to the person themselves, and to whichever family members they chose when they signed up. The alert says in plain English what the message is trying to do and why we think it is a scam. Nothing is deleted or hidden; the decision stays with the family.
Does a real person ever look at it?
Yes. Serious detections are reviewed by a working security architect before anyone is alarmed, and when you call or write, a person answers. The software does the reading; a human owns the judgment.
What does Family Sentinel miss?
We publish this rather than hide it. We watch email, so a scam that arrives entirely by phone call, text message, or social media is outside what we see — though the email that usually precedes or follows it is not. Our Scam Library marks every scam type as fully detected, partly detected, or not yet, and we keep that list honest.