Protection
without surveillance.
Guardii warns people about grooming, sextortion, scams and abuse in their messages. It runs on their own phone, like Face ID, so nobody else reads them.
Schools, clubs, employers and police get the reports people choose to send, ready to act on.
backedLaw Enforcement
Reports people choose to share, packaged so investigators can act after a person has reviewed them.
OpenSchools & Districts
Reports students and families choose to share, in your safeguarding lead's hands.
OpenAthletes & Clubs
An app on the player's own phone. The club sees nothing unless a player sends it.
OpenPlatforms & Enterprise
Protection your people adopt, and no new window into anyone's messages for you.
OpenWatch it catch grooming as it builds.
Grooming starts small: a compliment, a question about age, a push to move to another app. No single message is a bad word, so keyword filters miss it. Guardii reads how the conversation builds.
Conversation simulated from real grooming case data. Loops every 25s.
Awaiting signal…
- 01
Messages stay on the phone
Encrypted there, with a key only you hold.
- 02
The AI checks for patterns, right there
Like Face ID, it works on the phone and sends nothing away to do it.
- 03
Only a warning leaves, never the words
The person it protects sees it first.
Fits the way your team already works.
Families get it on their phones. Your team gets the reports people choose to send, set up around your policies and delivered where you already work.
AssistantShared reports onlyTwo reports were shared with the school this week, one already resolved. You see only what the student or family chose to share, never their messages.
GuardThe harm moved into private messages.
Grooming, sextortion and bullying happen in DMs and group chats now. No parent, teacher or moderator is in the room, and most kids never tell anyone. Guardii spots the warning signs on the phone and tells the person first, without anyone reading their messages.
of cyberbullied children were last targeted in a group chat (eSafety, 2024–25)
of parents knew their child had been treated hurtfully online (eSafety)
of children aged 10–17 have been told online to hurt or kill themselves (eSafety)
grooming-related reports to NCMEC's CyberTipline in 2025 (NCMEC)
The phone spots the warning signs.The person decides what happens.
Guardii spots the warning signs on the phone. The person it protects is warned first and decides what happens next. Nothing goes further without their consent.
It runs on the phone
Like Face ID
Guardii checks the apps the person connects, right there on their phone.
It reads the build-up
Not single words
Grooming, sextortion and harassment build over time. Guardii spots the build-up, not just bad words.
How serious, how sure
Two separate answers
Every warning says how serious it looks and how confident Guardii is, separately.
The person is warned first
Then a parent or chosen contact
The person being protected sees the warning before, or as, anyone else does.
They decide what happens
With a person in the loop
Evidence goes further only with their consent, and to police only after a person has reviewed it.
Private by design.
People take Guardii up because it does not watch them. That is what makes it work.
Messages stay on the phone
Guardii checks them there. We never receive them, so there is nothing for us or anyone else to browse.
People know it is there
Nobody is covered behind their back. They choose what is covered and who hears about it.
Warnings, not transcripts
What leaves the phone is a warning, or evidence the person chooses to send. Never a feed of their messages.
English and Arabic, dialects included.Checked by native speakers.
Guardii works in English and Arabic today. Every language we add is checked by native speakers before it goes live, which is the part that takes time.
English
Slang, abbreviations and emoji included.
Arabic
Modern Standard and Gulf dialects.
Mixed messages
English and Arabic in the same chat, and Arabic typed in Latin letters.