Could this be true?
AI detection that shows its work. Paste text, drop an image — we surface the artifacts. ELA, FFT spectrum, channel decomposition, noise residuals, C2PA provenance. Not a black-box score. The evidence.
How it works
Six signals, one verdict
Every check runs an ensemble of forensic techniques. Each one produces a visualization you can inspect — so you always know why something was flagged.
Error Level Analysis
ELARe-saves the image at known JPEG quality and measures compression delta. Splices and edits glow.
FFT Spectrum
FFTFourier transform reveals periodic artifacts left by GANs and diffusion samplers — invisible to the eye, loud in frequency.
Channel Decomposition
RGB / YCbCrSplits the image into color channels. AI generators leak color statistics that don't match real cameras.
Noise Residual
PRNUReal cameras leave a unique sensor-noise fingerprint. AI images either lack one or have a learned, repetitive one.
Saliency Heatmap
Grad-CAMWhere is the detector model paying attention? Heatmap shows the regions driving the verdict.
C2PA Provenance
CryptographicReads tamper-evident signatures from Adobe, OpenAI, Sony, BBC content. When present, this beats every other signal.
Why we're different
Most detectors give you a number. We give you the receipts.
Originality, GPTZero, Hive — every major AI detector outputs a single confidence score: “87% AI.” You're asked to trust a black box. We don't do that.
We expose the entire forensic pipeline. Every signal is visualized, every decision is auditable. Combined evidence beats any single classifier — and harder to evade than one.
Multi-signal ensemble
Six forensic checks run in parallel. One fails, others catch it.
Visual evidence
Every signal renders to a heatmap, spectrum, or channel split you can inspect.
C2PA-aware
Reads cryptographic provenance from Adobe, OpenAI, Sony, and BBC content.
Cheap, fast, private
Most checks run in your browser. Nothing uploaded unless you ask.
Field notes
How to tell what's real
- guide
The Tells of an AI Image: A Visual Forensic Walkthrough
What artifacts give away AI-generated images? A guided tour through ELA, FFT spectra, channel decomposition, and the diffusion-model fingerprints you can see with your own eyes.
11 min read
- guide
How to Spot AI-Written Text: 9 Signs That Actually Work
A field guide to recognizing AI-generated writing — from burstiness and perplexity to the telltale tics of GPT, Claude, and Gemini. Practical examples, not theory.
8 min read
- explainer
What Is C2PA, and Why It Will Quietly End AI Detection
Cryptographic content provenance is rolling out across Adobe, OpenAI, Sony, and the BBC. Here's how it works, who's signing what, and why detection tools should embrace it instead of fight it.
9 min read
Stop guessing. Start verifying.
Drop a paragraph. Drop an image. See the evidence in 3 seconds.