Truvizy — AI Content Scanner

Truvizy protects you from scams, deepfakes, and AI-manipulated content as you browse the web. Automatic, on-device, and free to start. 🛡️ REAL-TIME VIDEO CALL PROTECTION Detects deepfake video during browser-based calls on Google Meet, Zoom, Teams, WhatsApp Web, and Discord. A live verdict badge appears on each participant. No setup, no screen sharing, completely passive. 🔍 AUTOMATIC WEBSITE SCANNING Every page you visit gets silently analyzed for phishing, scam patterns, fake authority claims, urgency tactics, and hidden content. If a site looks dangerous a clear warning overlay appears with specific threat details. 🎬 VIDEO SCAN BADGES Scan badges appear on videos across YouTube, TikTok, Instagram, and X. Click any badge to instantly check for deepfakes and scam content. 🧠 ON-DEVICE AI Built-in AI runs entirely in your browser for the quick scan. No video frames, audio, or page content are ever sent to external servers for that step. Your browsing stays private. ⚡ KEY FEATURES • Passive deepfake detection on video calls, zero friction • Multi-signal website scam detection (URL plus page content) • Video scan badges on YouTube, TikTok, Instagram, X • Right-click context menu to scan any link, image, or video • Browser notifications for high-risk detections • Scan history (last 50 scans) • Works with your Truvizy account for AI-powered deep scans • Dark mode interface 🔒 PRIVACY FIRST Quick scans run on your device. No browsing history is collected or transmitted. You control all permissions and can revoke them anytime. Deep scans (optional, account-required) send only the specific link or media you choose to verify. Works with Google Chrome, Brave, Microsoft Edge, and all Chromium browsers. Free to use, no account required for basic protection. Sign in with your Truvizy account for AI-powered deep scans. What's new / Release notes v1.4.0 🎯 ACCURACY FIRST (NEW in v1.4) Verdicts are now more honest about how confident the analysis actually is. When a scan runs with reduced data the result clearly says so instead of returning a misleading score. If the analysis cannot be completed at all, you see a neutral "Unable to verify" outcome and no credits are charged. Honesty over false confidence. 📐 TOOLTIP NEVER GETS CUT OFF (NEW in v1.4) The verdict details panel now adapts to the space your screen has and scrolls inside itself when needed. Tall feeds on X, TikTok, and Reddit no longer clip the panel off the bottom of the screen. 📬 EMAIL BADGES NOW CONSISTENT (NEW in v1.4) Gmail inbox row badges and the in-email warning banner used to disagree about the same email. Now they agree. Open an email and the inbox row updates to reflect the full analysis. The banner is in English and uses the word "Risk" so a 0 reads as safe at a glance instead of confusingly low. 📦 SMALLER, FASTER (NEW in v1.4) Extension package dropped from 15 MB to 3.2 MB for a faster install. A clear progress label tells you what is happening on first use. Lower CPU on social feed pages, less battery, same coverage. 🔐 SECURITY UPDATE (v1.4) Patched a critical security vulnerability in a third-party library. No new permissions, no new data flows. v1.3.1 🎙️ VOICE DEEPFAKE DETECTION Real-time detection of cloned and AI-generated voices on the tab you are watching. Completely passive. A badge appears when something sounds off. Deep voice analysis is opt-in and uses zero credits until you tap Verify. 🧠 ON-DEVICE DEEPFAKE MODEL AI deepfake detection runs locally in your browser for instant image and video verdicts. No content leaves your device for the quick scan. 📱 REDDIT AND GMAIL SUPPORT Full scanning pipeline on Reddit posts and images, plus automatic badges on suspicious Gmail messages. 🔗 LINK PREVIEW SCANNER Scam-URL analysis on link preview cards across all supported feeds. Shows the source domain and a trust verdict inline with the post. New features in v1.3 • Voice deepfake detection on the active tab (passive, real-time) • On-device AI for instant quick scans • Two-phase Deep Scan with cloud confirmation (account required) • Reddit and Gmail platform support • Link preview scanner with scam-URL analysis • AI content detection from Grok, Ideogram, and more • Image scanning on every supported platform • Redesigned tabbed evidence panel with smart positioning Reliability and UX fixes • Click-to-scan on social feeds, no auto-scanning, no wasted credits • Fixed TikTok search returning the same verdict for different videos • Fixed missing badge in the TikTok video player modal • Fixed an AI model loading error under strict browser security policies • Smaller package, faster load, hard cap against duplicate badges