Scanner
Is it a scam? Paste it and find out.
Check a suspicious link, text message or email, or see whether a password has leaked. You get a verdict, how sure the AI is and the reason in plain language, so you know what to do next. 3 free AI checks a day, no account, no card.
3 free AI checks a day, no account, no card. What you paste is hashed before anything is kept, and passwords are hashed on your device.
By running a check you agree to our Terms and Privacy Policy. Cloudflare Turnstile checks you’re not a bot.
AI verdicts can be wrong. Treat one as a second opinion, not a guarantee.
Hashed first
Your input is hashed with SHA-256 before anything is kept. We don't store the full link or message. If a link is rated a scam with 80%+ confidence, we keep the website's name, never linked to you. Passwords are hashed on your device, and the password itself never leaves it.
Reasoned, not blacklisted
The verdict weighs the domain, the wording and the pressure tactics together, instead of matching a list.
Explained in plain language
Every call comes with a confidence level and named reasons you can check and disagree with.
Check links and text right from your browser. The extension is free: sign in with Google or Microsoft for 3 AI checks a day.
What happens to what you paste
- You paste A link or message
- Hashed at our edge Turned into a one-way code
- AI reads it in memory Nothing written down
- Verdict comes back Verdict, confidence, reason
- What we keep Site name only if a scam 80%+
Step 1 of 5 · You paste
https://secure.paypa1-login.com/verify?id=48213&email=you@example.com
You paste a link, text message or email you are unsure about.
- 1. You paste: You paste a link, text message or email you are unsure about.
- 2. Hashed at our edge: Our edge server turns it into a one-way code (SHA-256). That code, not your text, is what we keep. The code doesn't contain your text.
- 3. AI reads it in memory: The AI reads the text in memory to judge it. We don't save it. Our AI provider, Anthropic, may keep it for up to 30 days for safety review and doesn't train on it.
- 4. Verdict comes back: You get a verdict, how sure the AI is (0–100%), and the reasons in plain language.
- 5. What we keep: Always: the one-way code, the verdict and the AI's reason. Only if the AI rates it a scam with 80%+ confidence: the website's name, never the full link and never your name or account. To stop abuse, each check also logs a scrambled (salted) form of your network address for up to 2 years.