Independent image verification
Was this image made by AI?
Upload a picture and get a clear verdict backed by evidence. Every scan checks the data hidden inside the file, verifies Content Credentials, runs a detector trained on AI images, and adds a close visual review of the pixels.
- Content Credentials verification
- Trained AI-image detector
- Visual review
- Camera and edit checks
Your image
No file loadedVerification report
How every verdict is reached
1 · File evidence
Runs in your browser. It reads camera settings and AI labels, verifies Content Credentials (the signed records ChatGPT, Adobe Firefly and some cameras attach), compares the camera's saved preview with the picture, and flags files whose type doesn't match their name.
2 · Trained detector
A model trained on large sets of real and AI-generated images scores how likely the picture is AI-made. When it runs, its score leads the verdict.
3 · Visual review
An AI vision model checks anatomy, text, patterns, lighting, reflections and texture, and lists what it found. Its percentage is an estimate, not a measured accuracy.
4 · Verdict rules
If the file's own data says an AI tool made it, that decides it. Otherwise 70% or higher is likely AI and 30% or lower is likely not. When the checks disagree, or camera data contradicts them, the verdict says mixed signals.
No detector is always right. Files saved from social apps or screenshots lose their hidden data, edited photos can look artificial, and the newest AI images can look flawless. The error level map in each report is a tool for a closer look, not a verdict. Treat the result as strong evidence, not proof.