What Error Level Analysis Actually Tells You
ELA is the most misread tool in image forensics. Here is what the colours mean, what they don't, and how to avoid the classic false positive.
Read articleBlog
Field notes on image forensics
Practical guides to verification — how the techniques work, where they fail, and how to read a result responsibly.
Seven Signals That an Image Came From a Diffusion Model
Generators have fixed most of the obvious tells. These are the artefacts that survive into 2026, and the ones that no longer work.
Read article
Why We Refuse to Upload Your Images
Every verification service asks you to hand over the file. We built the opposite, and the architecture is the whole argument.
Read article
How to Read an Authenticity Score Without Fooling Yourself
A number between 0 and 100 invites false confidence. Here is how to interpret the score, the bands, and the layer breakdown responsibly.
Read article
Sensor Noise Is a Fingerprint. Here Is How to Read It
Real photographs are noisy in a very specific way. Understanding that specificity is the foundation of generative-artefact detection.
Read article
Where Deepfakes Still Break in 2026
Face-swap quality has improved enormously. These are the boundaries, blend seams and temporal artefacts that still betray it.
Read article
A Practical Verification Workflow for Newsrooms
Forensics is step four, not step one. The workflow that catches the most fakes starts somewhere else entirely.
Read article
What JPEG Compression Leaves Behind
Every save writes a signature into an image. Reading those signatures reveals how many times a file has been through an editor.
Read articleWhat EXIF Metadata Reveals — And Why Its Absence Proves Nothing
Camera metadata is rich, useful and trivially forged. A guide to using it without over-trusting it.
Read articleThe Complete Guide to False Positives in Image Forensics
Authentic photographs fail forensic tests constantly. Knowing exactly why is the difference between analysis and accusation.
Read articleContent Credentials: Provenance Beats Detection
Detection is a losing arms race. Cryptographic provenance changes the question, and it is arriving faster than most people realise.
Read articleAn Image Forensics Glossary
The vocabulary you need to read forensic output, research papers and expert reports without getting lost.
Read article