AI Photo Fake Detector analyzes a photo’s forensic signals and summarizes them in plain language. It combines five independent checks—Error Level Analysis, noise consistency, stylization artifacts, JPEG blockiness, metadata red flags and much more—then outputs a verdict with confidence plus an ELA visualization and human-readable notes.
Use it to triage images from social, messages, or your own camera roll. Nothing leaves your device unless you share a result.

Multi-signal analysis. One view, five checks, less guesswork.
Confidence and transparency. See the score and the evidence.
Clear ELA heatmap. Bright regions can indicate re-compression or edits.
Metadata notes that matter. Camera make/model, timestamps, JFIF, software tags.
Private by design. Runs offline. No accounts. No ads. No tracking.
Fast. Results in seconds on modern iPhones and iPads.

ELA anomaly: re-compression differences that can reveal edits.
Noise heterogeneity: inconsistent sensor noise across regions.
Stylization artifacts: AI/filters that leave style fingerprints.
JPEG blockiness: abnormal blocking beyond expected compression.
Metadata red flags: missing camera fields, software edits, JFIF export, etc.

Pick a photo.
Review verdict, confidence, scores, and ELA map.
Read metadata notes and decide with context.
Forensics estimate likelihood. They are not proof. Use context and corroborating evidence.
Dynamic Type, VoiceOver labels, high-contrast friendly.