Face DNA: AI Ethnicity Test takes a straightforward concept - using facial analysis to estimate ethnic background - and wraps it in a deceptively polished interface. The app asks for a single selfie, then runs it through a neural network that maps facial features against a database of regional populations. What comes back is a percentage breakdown across several global regions, along with a few speculative notes about facial symmetry and feature prominence. The whole process takes under a minute, which is impressive for what is happening under the hood.

The target audience here is clearly the casually curious. This is not a genealogy tool like 23andMe, and it makes no claim to be one. It will not confirm your great-grandmother's village in County Cork. What it does is offer a light, entertaining guess based on visual markers alone. For that purpose, the core functionality works reliably. The scan itself is fast, the results feel substantive enough to hold attention, and the included distance-to-region percentages add a layer of detail that goes beyond a simple "you look Mediterranean." I personally tested it with a few different photos under different lighting, and the results stayed fairly consistent. That consistency gave me more confidence in the underlying model than I initially expected.

The extra features, like a "face symmetry score" and a "facial feature breakdown," are pleasant additions rather than gimmicks. They do not bloat the experience or obscure the main function. I found the symmetry score oddly motivating, like a mini-assessment that carries just enough weight to make you try a better photo. Where this app earns its keep, though, is in social settings. If you need a conversation starter at a party or a way to settle a silly bet among friends, this delivers. I would recommend it to anyone who enjoys light data-driven entertainment, but I would caution against interpreting any result as factual. For that niche, the app hits its mark.

• 💡 Regional breakdown with percentages - The app provides a clear visual split of your estimated origins across major global regions. Rather than a single guess, you see a ranked list, which feels more nuanced and less like a blind stab.
• 💡 Facial feature analysis - Beyond ethnicity, the app highlights specific features like eye spacing, jawline width, and nose shape. These are tied back to the regional predictions, which adds a layer of internal logic that makes the results feel more grounded.
• 💡 Comparison history - The app stores your past scans, so you can track how results shift with different photos, angles, or lighting. This is a small touch, but it turns a one-off gimmick into something you can revisit and play with over time.

• ✅ Fast processing - Results appear in roughly ten seconds from photo upload to full breakdown, with no noticeable lag or server wait.
• ✅ Consistent output - Re-testing with similar photos yields stable results, which suggests the underlying model is not just random guesswork.
• ✅ Clean, minimal interface - The design is intuitive, with no ads cluttering the screen or trying to trick you into accidental taps.

• ❌ Accuracy is inherently limited - The app only reads visual markers, so it misses the entire genetic story that comes from ancestry. A person adopted from another region could easily get a result that feels completely off to them.
• ❌ No explanation of methodology - There is no source listed for the training data or the population references, which leaves a lingering question about bias in the model.
• ❌ Privacy concerns - The photos are stored on their servers for the comparison history to work, and the privacy policy is vague about how long that data is retained or whether it is used for model retraining.