AiTruthScan

About

Detailed Description
🤖 Is it real or AI? Find out in seconds.

AiTruthScan scans images, videos, text, and audio for AI generation using multiple LLMs simultaneously — then combines their verdicts into a single consensus score you can trust more than any one model alone.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

⚡ HOW IT WORKS

Right-click any image, video, or text on the web to scan it. A floating toolbar appears when you hover over media. Results show three color-coded scores rated 0–100:

🔴 AI Generated — probability content is AI-made
🟠 Digitally Edited — probability content was manipulated
🟢 Authentic — probability content is real

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🧠 MULTI-LLM CONSENSUS ENGINE

Instead of trusting a single AI, AiTruthScan queries up to 4 providers in parallel and uses a weighted consensus algorithm for more reliable results:

✦ OpenAI (o3 / GPT)
✦ Google Gemini (Flash / Pro)
✦ Anthropic Claude (Haiku / Opus)
✦ xAI Grok (Fast / Grok-4)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🔍 WHAT YOU CAN SCAN

🖼️ Images — AI art, deepfakes, manipulated photos, AI portraits
🎬 Videos — frame-by-frame analysis with temporal consistency checks
📝 Text — AI-written articles, essays, social media posts
🔊 Audio — AI-generated speech and voice clones
📁 Local Files — drag & drop images, PDFs, DOCX, audio & video files

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✅ BUILT-IN FACT CHECKER

Claims found in images or text are automatically verified against web sources. See truth scores, source links, and search summaries alongside AI detection results. Works on screenshots of social media posts, news headlines, and more.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📤 SHAREABLE REPORTS

One-click share button generates a branded report card (PNG) with:
🖼️ Scanned media thumbnail with watermark
📊 Verdict banner + three-axis score bars
🔬 Synthetic indicators & forensic observations
📋 Fact-check results (when applicable)

Ready to paste into social media, group chats, or articles.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🎯 TWO QUALITY TIERS

⚡ Standard ($0.02/scan) — fast, affordable models for everyday use
🔬 High-End ($0.10/scan) — top-tier reasoning models for difficult or disputed content

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🔒 PRIVACY FIRST

🔑 Bring Your Own Keys — uses YOUR API keys, not ours
🔐 All keys encrypted with AES-GCM before storage
🚫 No external servers · No data collection · No accounts
💻 Everything runs locally in your browser

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🛠️ FEATURES AT A GLANCE

✦ Multi-model consensus with dissenter detection
✦ Forensic rendering evidence weighting
✦ Temporal persistence analysis for video compositing
✦ NSFW classification
✦ Content tagging (clickbait, engagement farming, verified, etc.)
✦ Source reputation tracking across scans
✦ Scan history with full detail review
✦ Keyboard shortcut: Alt+Shift+S
✦ Works on any website

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

👥 WHO IS THIS FOR?

📰 Journalists verifying sources
👨‍👩‍👧 Parents checking what kids are sharing
🔬 Researchers studying AI-generated content
🌐 Anyone tired of AI slop flooding the internet

⚠️ Don't take results at 100% face value — but it gives you something to think about.