A 7-Point Checklist for Spotting Deepfakes
Detectors like TruthLens AI exist for a reason — but a quick visual audit catches a lot before you ever upload. Walk through these in order.
1. Look at the eyes
Deepfaked eyes often have inconsistent reflections (each eye should reflect the same light source) and unnaturally crisp irises.
2. Check the ears and hairline
Face-swap models struggle with the boundary between face and head. Look for blurred or melted ears and a hairline that doesn't match the head shape.
3. Teeth and jewelry
Individual teeth often blur into one shape; earrings and necklaces frequently render asymmetrically.
4. Hands
Even in 2026, hands are a weak spot. Count fingers, look for warped knuckles or rings that twist into the skin.
5. Background text
Any signage, books, or screens in the background is a great tell — generated text rarely spells anything coherent.
6. Lighting direction
All subjects should be lit from the same direction. A face that looks lit from the left in front of a sunset that's behind them is suspicious.
7. Reverse image search
If the photo claims to depict a real event, run it through Google Lens or TinEye. Recycled stock photos and prior versions often surface immediately.
When in doubt, drop the image into the TruthLens detector for a confidence score and a list of the specific artifacts the model found.
Keep reading
How AI Image Detection Actually Works in 2026
A plain-English tour of the techniques behind modern AI image detectors — from diffusion-noise fingerprinting to identity-consistency checks for deepfakes.
Adding AI Image Detection to Your App in 5 Minutes
A short guide to integrating the TruthLens AI REST API for moderation, content trust, and journalism workflows.