Local simulation scanner
See what an AI model may see.
Inspect structural exposure and get practical protection guidance. Your image never leaves this device.
Drop an image into the scanner
JPG, PNG, or WebP · maximum 20 MB
No upload. No storage. No account.
Vulnerability & Security Report
Your local structural analysis will appear here.
This privacy-preserving simulation evaluates image structure and sharing risk; it does not detect whether an image is fake and cannot predict or prevent misuse.
About the AI Image Vulnerability Scanner
What it does
The AI Image Vulnerability Scanner inspects the structural fingerprint of any photo you choose — directly in your browser. It analyses dimensions, color depth, embedded EXIF metadata, GPS coordinates, device identifiers, and the visual features that generative AI models typically exploit to clone a face. Nothing is uploaded; the moment you close the tab, the analysis vanishes.
Why privacy matters here
Most victims of deepfake abuse never realise that the photo they posted years ago still leaks GPS coordinates, camera serial numbers, or facial geometry that can be harvested by scraper bots. Knowing exactly which fields are exposed is the first step toward removing them. A vulnerability scan turns an invisible problem into an actionable checklist — and because every check happens on-device, no third party (including us) ever sees your image or your conclusions.
How it helps victims of deepfake abuse
Survivors of intimate-image abuse, doxxing, and AI face-cloning use the scanner to triage which photos must be wiped immediately, which can be safely re-shared after metadata stripping, and which should be replaced with protected versions. The report it produces can be attached to DMCA notices, EU AI Act complaints, or police evidence files as objective, reproducible proof of the risk the original photo carried.
Frequently asked questions
- Is my photo uploaded anywhere during the scan?
- No. The entire analysis runs inside your browser using local JavaScript and the Canvas API. Your image never leaves your device, and we never receive a copy or a fingerprint of it.
- What kinds of vulnerabilities are detected?
- EXIF metadata exposure (GPS, camera model, timestamps), high-resolution facial detail used by AI cloning models, and structural cues that reverse-image search engines rely on to track a photo across the web.
- Will the scan remove the vulnerabilities for me?
- The scanner is read-only. To actually strip metadata and apply a protective AI-shield watermark, use the 'Lock My Photos from AI' tool, which also runs 100% in-browser.
- Can I trust the scan results in a legal complaint?
- Yes. The scanner reports objective, reproducible facts about the file (hash, metadata fields, dimensions). You or a forensic expert can re-run the same checks independently to verify the findings.