Anything downloaded from Instagram has already been stripped of metadata and re-compressed twice. That kills the easy checks and leaves forensic analysis of the pixels — which still works, just with less margin.
Instagram re-encodes every upload and removes metadata. Some evidence goes with it; plenty does not.
Camera, timestamps, GPS — stripped on upload. Their absence tells you nothing at all here, because it is true of every image on the platform.
Frequency profile, texture coherence and noise distribution survive re-compression well enough for the neural models to work with.
Generated portraits keep their tell-tale alignment regardless of how many times the file is re-saved.
Compression does not invent a coherent room. If the setting behind the person is incoherent, that survives too.
Heavy filters and beauty modes overwrite texture, which is exactly the evidence we rely on. Expect lower confidence on them.
Each re-upload compresses again. A picture that has been through five accounts is far harder to judge than a first-hand post.
Full resolution, straight from the source. Screenshots and re-saved copies destroy most of the evidence we measure.
Three neural networks look at the image as a whole. Nine forensic methods examine it as a signal: frequency (FFT) analysis, noise pattern, pixel-level statistics, EXIF metadata, Error Level Analysis, texture coherence, colour distribution, Benford's Law and edge detection.
Every method reports separately, so you can tell whether the verdict rests on one weak signal or on nine that agree.
For images from anywhere else, start with our AI image detector — it covers any image type and runs the same twelve methods.
Most detectors give you one number from one model. We run twelve independent analyzers and weight them — three neural networks carry the majority, but nine forensic methods examine the file as a signal and can overrule them. Here is the exact weighting we use in production:
| Method | Weight in final score |
|---|---|
| Neural network #2 | 25% |
| Neural network #1 | 15% |
| Neural network #3 | 15% |
| EXIF metadata | 10% |
| Frequency (FFT) | 6% |
| Benford's Law | 6% |
| Noise pattern | 5% |
| Error Level Analysis | 4% |
| Texture coherence | 4% |
| Edge detection | 4% |
| Pixel statistics | 3% |
| Colour distribution | 3% |
These are our own numbers, not an industry estimate — every check run through FakeSec since launch. The split is closer than most people expect: roughly four in ten uploads really do turn out to be generated.
On our internal benchmark of 240,000 images (120,000 AI-generated, 120,000 real) the combined score reaches up to 98% on unedited, full-resolution files. That number drops, sometimes sharply, in four situations — and you should know them before trusting any result:
No detector on the market is 100% accurate — ours included. Treat a high score as strong evidence, not as proof. When a decision matters, look at which of the 12 methods agreed.
Confidence is genuinely lower on social downloads, and we would rather flag that than dress it up. If you can obtain the original file rather than a screenshot of a repost, the answer is much firmer.
Create a free account and get 3 checks. Full breakdown across all 12 methods, no card required.