ProofGuard logo ProofGuard
Log in Sign up

Detect AI
Content

100% Automatically and Free

Supported
Images (PNG/JPG/WebP), PDFs, and plain text.
Local history
Your last 15 uploads are saved only in your browser.
How it works
We analyze metadata, text signals, and visual forensics to estimate likelihood.

Recent uploads

View all →
How ProofGuard works
Upload or drop
Add an image, PDF, or text. We generate an in-page preview where possible to show what’s being analyzed.
Analyze multiple signals
We combine metadata checks, text cues, and visual forensics (like edge density and ELA) to estimate likelihood.
Score + explanation
You’ll get a confidence score with a breakdown of top contributing aspects and any notable indicators we found.
Signals we analyze
Metadata indicators

EXIF presence, camera/codec traces, JPEG Q-tables, and missing/edited fields.

OCR AI terms

Keyword hits and phrasing patterns often associated with model output.

Edge/texture cues

Edge density, Laplacian variance, blockiness—flat or overly smooth regions can be suspicious.

ELA and compression

Error level analysis and low-ELA regions that may indicate synthetic composition.

Color/brightness stats

Distribution and uniqueness checks (e.g., chroma/luma ratios, brightness/saturation moments).

Text repetition/TTR

Repetition, token variety, and sentence length distribution for text-based content.

What to expect

The confidence score reflects the strength of observed signals, not a definitive ground truth. Some benign workflows (like exports or aggressive compression) can look AI-like, while sophisticated generations may evade detection. Always combine results with human judgment and source context.

Privacy & local history

Your last 15 uploads are stored only in your browser for convenience (see Uploads). For production deployments, set storage/retention policies to match your organization’s requirements.

For developers
Use the HTTP API Read the docs Integrations & Plugins