This AI Image Detector Checks a Global Database of Debunked Pictures

AI image detectors are notoriously imperfect, giving an estimated score rather than concrete answers. But a new tool called Image Whisperer checks to see if a picture has been debunked elsewhere — acting more like a human researcher.

Henk van Ess of Digital Digging compares current AI image detectors to a health inspector with a checklist examining a kitchen and finding a rat in the kitchen wearing a tiny chef’s hat.

“The checklist didn’t have a question about rats in chef’s hats,” van Ess says. “Mathematically, the restaurant only loses a few points for ‘pest control.’ But any human with functioning eyes would say: I don’t care what the checklist says – that rat is running the kitchen.”

While Image Whisperer v1.0 use “multiple AI-detection models” to analyze images, it will also run a reverse image search to identify whether the image has been debunked or verified elsewhere — much like a journalist would when trying to find out whether a piece of media is real or not.

The multi-layer detection system also uses “four different AI models” to examine an image for visual anomalies humans would notice, including “melted faces, impossible physics, inconsistent lighting, and architectural impossibilities.”

But instead of giving a score like other AI image detectors, Image Whisperer delivers a color-coded verdict with a “clear explanation of why.”

“ImageWhisperer doesn’t just output a number,” van Ess says. “It weighs evidence from multiple detection systems, applies LLM judgment, and delivers a color-coded verdict with a clear explanation of why.”

Red: AI-Generated. Strong evidence of AI generation. Multiple detection systems agree. Critical indicators found.

Orange: Uncertain. Mixed signals from detection systems. Some concerning indicators but not conclusive. Requires human review.

Green: Likely Real. Passes most tests. No critical failures. Consistent noise patterns and physics. Consistent with real photography.

Blue: Human Review. Image found in news sources. Conflicting reports exist. Human verification essential to determine authenticity.

Today I launched Image Whisperer v1.0 — out of beta! Upload a suspicious photo and it instantly checks against a global database of debunked images, updated daily. If your photo matches a known fake, it shows you the debunk. Try it: https://t.co/9QyrkIUhk3 (1/3) pic.twitter.com/tj49FHiBZn

— 𝚑𝚎𝚗𝚔 𝚟𝚊𝚗 𝚎𝚜𝚜 (@henkvaness) February 10, 2026

While it’s a welcome tool to help aid the discovery of fake images, van Ess warns it “cannot guarantee 100% accuracy on any image.”

“This is a helper tool, not a judge,” he adds. “No automated system can replace critical thinking and journalistic verification. It’s best used as a first-pass filter, not a final arbiter.”

In PetaPixel’s tests, Image Whisperer performed well. Try for yourself here.

Image credits: Header photo licensed via Depositphotos.