SIGNET
Sign inGet started

For everyone

Your face and voice are the new credit card.Treat them like one.

AI-cloned voices are scamming families. Deepfaked faces are running romance scams. Your photos are training the next model. Signet catches it and takes it down.

Start protection See pricing

What you're actually up against

"Grandparent" voice-clone scams
A 10-second voice sample from your TikTok is enough to call your parents at 2 AM claiming you are in a car accident and need bail money. The FTC logged $2.7B in imposter-scam losses in 2023, and voice-cloning attacks are now the fastest-growing category.
Hijacked or impersonated accounts on Facebook, Instagram, TikTok, dating apps
An attacker takes over your account (or builds a clone with your photos) and starts scamming your network — sponsorship fraud, romance scams, "I need bail money" messages to your followers. Meta's automated impersonation flow trusts whichever account has more history — when the attacker has been operating for months and you report from a fresh account, the system sides with them. Real, documented, common.
Deepfake nudes uploaded to revenge sites
AI nudification tools turn any clothed selfie into a fake nude in seconds. The fake gets uploaded to "leak" sites that monetize the SEO. Most sites WILL take it down on a DMCA notice — if you can find every copy.
Your photos scraped into AI training sets
LAION-5B, Common Crawl image dumps, and dozens of less-public datasets vacuumed billions of personal photos off the open web with no opt-in. The next generation of AI models is being trained on those right now.

How Signet protects you

Four pillars. The first three are included on every paid tier. AI training opt-outs require Pro.

Verified identity, encrypted at rest
One-time government-ID + selfie verifies you are protecting yourself, not impersonating someone. The biometric data is encrypted with AES-256-GCM and revocable any time — see the consent details before you commit anything.
Reverse face search across the open web
Monthly scans (Shield) or weekly (Pro) hunt for face matches on Reddit, TikTok, dating apps, leak sites, and image dumps. Each hit is shown to you with confidence score and screenshot for review.
Biometric-verified takedown filings
Click "take it down" on a detection. We file on your behalf — DMCA for copyright-based cases (stolen photos), platform Trust-and-Safety reports for impersonation. Every filing attaches your government-ID-verified identity record + biometric match + audit-logged chain of custody. That verification is what gets past Meta/TikTok/Instagram's automated layer to a human reviewer.
AI training opt-outs (Pro)
On Shield Pro and above, we submit opt-out requests to OpenAI, Anthropic, Stability AI, and Adobe Firefly on your behalf. Existing models cannot be untrained, but the bleed forward is plugged.

In practice

Parent of a teenager
Weekly scan surfaces an AI-edited "leaked" image of their child on a sketchy image board. Takedown filed within hours; image gone before it spreads to mirror sites.
Remote worker on LinkedIn
Reverse search catches a romance-scammer profile on a dating app using their LinkedIn headshot. DMCA + platform-trust-and-safety report filed; profile removed in 36 hours.
Privacy-conscious adult
Signs up for Shield Pro the moment AI training opt-outs become available. New training runs from the major providers exclude their face going forward, per the public opt-out commitments.

From early users

Signet is in private beta. Real testimonials replace these placeholders at GA.

“Quote from a happy beta user goes here.”
— Beta user #1
“Quote from a happy beta user goes here.”
— Beta user #2
“Quote from a happy beta user goes here.”
— Beta user #3
Recommended starting point
Shield
$12 / month
Catch impersonations as they appear.
  • Monthly automatic scans
  • Up to 10 authorized takedowns / year
  • Email alerts on every new detection

Active on social media? Consider Shield Pro ($29/mo) — weekly scans, unlimited takedowns, voice fingerprint monitoring, and AI training opt-outs.

Start with Shield Compare all tiers

Frequently asked

I am not famous — do I really need this?

Voice-cloning models need under a minute of audio. Face-swap tools need one good photo. Both are within reach of anyone with $20 and an evening. The targets shifted from "celebrities" to "anyone with a public social profile" two years ago.

How is this different from LifeLock?

LifeLock watches identifiers (SSN, credit-bureau records) for misuse. Signet watches biometrics (face + voice) for impersonation. Different attack surface — biometrics are now scammable in ways SSNs never were. Run both if you want full coverage.

What happens to my data if I cancel?

30-day cool-off window where you can restore the account. After that, all encrypted biometric data is purged within 7 days. Your audit log rows persist with your user_id nulled out — required for SOC 2 evidence retention, but it cannot be tied back to you.

Can I protect my whole family on one account?

V1 covers one person per account because biometric verification is per-person. Family plans (multi-identity per account, shared billing) ship in V2 by request — email support@signet.liteshare.org if this matters to your decision.

How accurate is the reverse face search?

Each detection comes with a confidence score (0.0-1.0). Anything above 0.7 is shown by default. False positives happen — that is why the takedown step is a deliberate "yes I confirm this is unauthorized" click, not automatic. We deliberately err toward "show too many" rather than miss.

Facebook ignored my impersonation report — can Signet help?

Yes, this is the case Signet is built for. Meta's automated system trusts whichever account has more history — when the attacker has been operating for months and you report from a fresh account, the system sides with them. Signet attaches institutional identity proof to every filing: government-ID-verified match, biometric attestation, append-only audit log. We don't override Meta's decisions, but the verification gets the report past automation to human review more often than a victim's own report does. The same evidence package also works for FTC IdentityTheft.gov filings and lawyer escalations.

Does Signet also scan the dark web?

Not in V1 — open-web monitoring only (TikTok, Instagram, Reddit, image-board mirrors, public dating apps, deepfake-leak sites that operate on the surface web). Dark-web monitoring is a different operational posture and ships as an add-on later.

Don't wait for the call from your mom.

Sign up + first scan in under 10 minutes. Cancel any time, deletion is permanent after a 30-day cool-off.

Get Shield See all plans
For everyone — Signet · Signet