Blog
AI Detection Guides & Research
Deep dives into how AI content detection works — ELA, FFT, C2PA, deepfakes and more.
Deepfake Job Interviews: How to Detect AI-Generated Video Candidates
Deepfake fraud in remote job interviews has surged 1,300% since 2022. This guide explains how it works, what warning signs to look for, and how forensic tools can help verify video authenticity.
Read articleDeepfake Romance Scams: How to Check if You're Talking to a Real Person
Romance scammers increasingly use AI-generated profile photos and deepfake videos to build fake relationships. This guide explains the warning signs and the forensic tools that can help you verify someone's identity online.
Read articleEXIF Metadata and Image Forensics: What Hidden Data Reveals
Every photograph taken with a real camera embeds invisible metadata — camera model, GPS coordinates, timestamp, lens settings. AI-generated images typically have none. Here is how to read EXIF data and what to look for.
Read articleFFT Spectrum Analysis: How Frequency Patterns Reveal AI-Generated Images
Fast Fourier Transform (FFT) analysis detects characteristic frequency signatures that distinguish AI-generated images from real photographs. Here is how it works and what to look for.
Read articleHow to Detect Midjourney Images: Forensic Tells and Detection Tools
Midjourney produces some of the most photorealistic AI images available. This guide covers the forensic signatures unique to Midjourney outputs and how to spot them with and without tools.
Read articleHow to Read ELA Heatmaps: A Plain-English Guide
Error Level Analysis (ELA) is one of the oldest image forensics tools — yet most explanations are written for academics. This guide shows you what to look for, what the colours mean, and what ELA cannot tell you.
Read articleMidjourney vs DALL-E: How to Tell AI Images Apart by Their Artifacts
Midjourney and DALL-E produce subtly different forensic signatures. Knowing the differences helps you identify not just that an image is AI-generated, but which model likely produced it.
Read articlePrivacy-First AI Image Verification: Why Zero-Retention Matters
Most AI detection tools store the images you upload. FakeRadar does not. Here is why zero-retention image analysis matters for journalists, researchers, lawyers and anyone handling sensitive visual content.
Read articleWhat is C2PA? Content Credentials Explained
C2PA is an open standard that attaches a verifiable provenance record to images, videos and audio — tracking where a file came from and what edits were made. Here is how it works, who supports it, and what it means for detecting AI-generated content.
Read articleWhy AI Detection Is Signal-Based — Not a Verdict
AI image detection tools do not produce a binary true/false verdict. They surface forensic signals. Understanding the difference makes you a better analyst — and a more sceptical consumer of detection results.
Read article