Guardian is in private beta. Sign up now to reserve your spot and upload content you want protected. Full monitoring activates by email invitation as we onboard in waves.

Continuous identity protection

Your face. Your voice. Your name. Protected.

Guardian protects your name, image, likeness, and voice from non-consensual imagery, impersonation, fabricated content, and unauthorized use. We monitor the surface and dark web, then prepare the takedown notices and evidence packages you submit to platforms.

  • No credit card required
  • TAKE IT DOWN Act notice generation
  • StopNCII hash submission
  • Biometric data never sold or shared
guardian.deepid.app/dashboard
Example dashboardv 2.4.1
Active Protection
All systems normal · continuous monitoring
Sources scanned
14,212 · last 7d
Active Jobs
NCII & Dark Web Discovery
608 forums · 42 leak sites
84%
Surface Web Identity Scan
14,212 sources
62%
Impersonation Monitoring
social platforms · news · domains
31%
NCII content discovered — leak forum — Notice ready to send
match 97%·TAKE IT DOWN Act notice prepared·StopNCII hash queued·48h response window starts at submission
Protected since Jan 2026·847 scans·3 resolved

Powered by Deep Media — Google partner, trusted by U.S. Government contractors and enterprise platforms worldwide.

TAKE IT DOWN Act48-hour
Deep Mediaenterprise
The threat

Your likeness is out there. Most people don't know until it's too late.

Non-consensual intimate imagery on sites you'd never think to look. Impersonation accounts wearing your face. AI-generated content attributed to you. Voice clones running scams in your name. The internet moves faster than any one person can monitor — Guardian watches for you.

> 1 in 5

adults has experienced non-consensual intimate imagery

Umbach et al., "Non-Consensual Synthetic Intimate Imagery: Prevalence, Attitudes, and Knowledge in 10 Countries" (ACM CHI 2024)

96–98%

of deepfake videos online depict non-consensual intimate imagery. 99% of victims are women.

Sources: Deeptrace Labs, "The State of Deepfakes" (2019); Home Security Heroes (2023); reconfirmed in multiple 2025 industry analyses.

48 hours

Maximum platform removal window mandated by the TAKE IT DOWN Act. Civil penalties of $53,088 per violation, enforced by the FTC effective May 19, 2026.

TAKE IT DOWN Act, S.146, 119th Congress; FTC compliance guidance

What sets us apart

Guardian goes where others don't.

Guardian protects against the full range of digital identity harms — non-consensual imagery, impersonation, fabricated content, voice clones, and unauthorized commercial use — across every layer of the internet, including the deep and dark web sources most consumer products don't monitor.

Layer 01

NCII & Dark Web Discovery

Guardian indexes breach forums, fringe platforms, ransomware leak sites, and underground image-sharing networks that consumer products typically don't touch.

Layer 02

Surface Web Identity Monitoring

Reverse image and video search against your facial and voice patterns across Google Images, Bing, Yandex, TinEye, social platforms, news sites, and high-risk impersonation domains.

Layer 03

Impersonation & Fabricated Content Detection

Detects impersonation accounts, fake bylines, domain spoofing, and synthetic audio and video attributed to you — including AI-generated speeches, voice clones, and fabricated interviews.

Layer 04

Facial Biometric Protection

Enroll once. Guardian uses encrypted mathematical embeddings — never raw images — to find your face across the surface, deep, and dark web.

Layer 05

Voice Biometric Protection

Voice cloning is the fastest-growing identity threat. Enroll your voice and Guardian monitors for unauthorized audio matching your vocal patterns.

Layer 06

Evidence Generation & Takedown Preparation

When Guardian finds a violation, it prepares the takedown package — notice, evidence, and platform-specific submission instructions. You review and submit; Guardian tracks the response through the 48-hour window.

Five steps to enroll

Set up once. Protection runs continuously.

No security background required. Once you finish enrolling, Guardian's monitoring jobs run on a continuous schedule — daily by default, with hourly cadence available for high-risk personas.

  1. Step 1/ 05

    Create your account

    Sign up with Google, magic link email, or another supported provider. No credit card required.

  2. Step 2/ 05

    Tell us about yourself

    Quick questionnaire about your role and protection priorities. Guardian uses this to route detection and alerts appropriately — a senator's threat profile is different from a creator's, and Guardian treats them differently from day one.

  3. Step 3/ 05

    Enroll your biometrics

    A 3-frame liveness capture (or three still photos) for your face. Read five short sentences (about 30 seconds) for your voice. Guardian converts these to encrypted mathematical embeddings — raw images and audio are deleted within 24 hours.

  4. Step 4/ 05

    Upload additional content

    Photos, videos, and audio you want protected. Up to 500 MB per file, 5 GB total per user during beta. Bulk roster upload available for agencies, professional associations, and political organizations.

  5. Step 5/ 05

    Guardian's monitoring jobs activate

    NCII discovery, surface web identity monitoring, impersonation detection, and fabricated-content monitoring run on your protection profile. Your first scan completes within 24 hours.

Audiences

Built for everyone. Especially those most at risk.

The product

Full visibility. Total control.

Your Guardian dashboard shows every scan, every finding, and every takedown in one place. Review evidence, send notices, escalate to your security or comms team, and track removals through to completion.

Example dashboard
MR
Findings · last 24h

Recent activity

allpendingresolved
NCII content — leak forum
resolved
match 98% · NCII · 2h ago

Intimate image of subject discovered on aggregator forum. TAKE IT DOWN Act notice and evidence prepared, hash submitted to StopNCII, content removed within 7 hours of your submission.

Voice clone audio — Telegram
ready to send
match 95% · Voice Match · 14m ago

AI-cloned audio attributed to subject promoting a fraudulent investment scheme. Notice and evidence package prepared. 48h response window starts at submission.

Scan throughput · 7d
14,212 sources
3 active jobs·Last scan: 14 minutes ago·All systems normal
Trust & compliance

Built with privacy at the center.

Guardian is protection, not surveillance. Every architectural decision starts with what we will never collect, never sell, and never share.

Biometric sovereignty

Your biometrics, your control

Stored as encrypted mathematical embeddings — never raw images or recordings. Delete anytime. BIPA, CCPA, and Washington My Health My Data Act compliant. Raw frames and audio are purged within 24 hours of embedding extraction.

Federal compliance

TAKE IT DOWN Act ready

Every notice is formatted to meet the Act's content and submission requirements — triggering the 48-hour platform response window once you submit. Notice templates are also available in DMCA, FTC unauthorized endorsement, and FEC electoral content report formats where applicable.

Integration

StopNCII integration

With your consent, Guardian submits hashes of verified NCII findings to the StopNCII database — the same system used by Facebook, Instagram, TikTok, Reddit, Snap, OnlyFans, Pornhub, Google Search, and Microsoft Bing to detect and remove non-consensual intimate imagery.

The internet doesn't forget.
Guardian makes sure your likeness isn't weaponized.

Free during beta. Biometric data never sold or shared.