About AIGuardian
Company Introduction
AIGuardian is an enterprise-grade visual AI forensics platform by YUMSHU focused on deepfake video detection and synthetic image verification. The platform helps security teams, journalists, and trust & safety reviewers analyze visual media with probability-based signals, explanation outputs, and workflow tools such as history, quota management, and account controls. Public site: yumshu.com.
How Detection Works (Methodology)
AIGuardian does not claim 100% accuracy. Outputs are probabilistic evidence for human review:
- Video: Frame-by-frame temporal consistency, facial landmark stability, and biological anomaly checks aimed at deepfake and synthetic video risk.
- Images: Frequency-domain (FFT) analysis and diffusion artifact checks for state-of-the-art generative image models (Midjourney, FLUX.1, DALL-E).
- Biometrics & Provenance: Photoplethysmography (PPG) pulse analysis and C2PA Content Credentials support.
- Human-in-the-loop: Help and product copy frame scores as signals—not automatic legal or editorial verdicts.
- Data handling: Analysis uses encrypted connections; uploaded files are deleted after analysis completes (see Privacy Policy).
Who We Serve
Self-serve buyers and teams that need visual forensics and media verification without long sales cycles. Typical customer types:
- Trust & safety / risk teams — deepfake video and synthetic media review.
- News media & journalists — breaking news photo and video authenticity verification.
- Cybersecurity & KYC teams — presentation attack defense and face-swap prevention.
- SMBs and creators — transparent Free / Weekly ($2.99) / Monthly ($7.99) / Quarterly ($19.99) / Annual ($49.99) plans on Pricing.
We are not optimized for air-gapped on-premise-only deployments or traditional classroom LMS grading suites.
Our Products
A visual forensics platform covering video, image, and deepfake detection:
- Video Detection: AI Video Detector analyzes frames and temporal sequences for synthetic video and face-swap risks.
- Image Scanning: AI Image Detector reviews synthetic images from common diffusion generators.
- Deepfake Playbook: Deepfake Detection Guide provides trust & safety procedures for synthetic media escalation and provenance verification.
- Gemini Multimodal Guide: Gemini Multimodal Forensics forensic review workflows for Google Imagen 3 and Veo media.
Product Updates
Dated milestones you can verify on the live site and sitemap (not anonymous claims):
- 2026-09 — Next-Gen Visual Forensics Engine: Calibrated video and image models against Sora 2, Kling, Veo 3, FLUX.1, and Midjourney v6.
- 2026-07 — Competitor alternative pages: Published self-serve comparison pages for Reality Defender, Sightengine, WasItAI, and Deepware at /alternatives.
- 2026-05 — Deepfake and Gemini playbooks: Released Deepfake Detection and Gemini Multimodal Forensics guides for trust & safety workflows.
- 2026-01 — Visual forensics suite launch: Unified deepfake video and synthetic image forensics under one AIGuardian account with published Free through Advanced quotas.
Core Values
- Innovation: Continuously evaluating detection models as generators change.
- Integrity: Ethical use framing and privacy-conscious handling of uploads.
- Accuracy: Focus on reducing false positives and improving explanation quality—not inventing certainty.
- User First: Self-serve quotas, clear pricing, and human-review guidance.
Contact Us
Questions, institutional pricing, or collaboration:
Email: support@yumshu.com
Address: Not publicly listed. Please contact us by email.