Qual rosto é REAL?
Rosto A
Rosto B
Você consegue distinguir qual rosto é real e qual foi gerado por IA?
This interactive game tests your ability to distinguish between genuine human photographs and AI-generated synthetic faces. Each round presents you with two portraits side-by-side - one is a real person captured by a camera, while the other is an artificial face created entirely by machine learning algorithms.
Your challenge is simple: identify which face is real. After making your selection, you'll receive immediate feedback showing whether your guess was correct. The game tracks your score, current streak, and overall accuracy percentage, allowing you to measure your performance over time.
As AI technology advances, the line between real and synthetic media becomes increasingly blurred. Understanding how to identify AI-generated content is becoming an essential digital literacy skill. This challenge not only entertains but educates, helping you develop a critical eye for detecting synthetic media in an age where deepfakes and AI-generated content are becoming commonplace.
Look for these common tells:
Important Note: As AI technology improves, these tells become less reliable. Modern AI face generators like Google's Imagen produce incredibly realistic results that can fool even trained professionals. The best approach is to look for multiple subtle clues rather than relying on a single indicator.
Train your eye to spot synthetic media - a crucial skill in the age of deepfakes and AI-generated content.
Most people score only 50-60% accuracy. Can you beat the odds and develop superhuman AI detection skills?
Monitor your accuracy, score, and streak to see how your skills improve with practice.
Quick rounds make it perfect for a break, while the challenge keeps you coming back to improve your score.
Understand the capabilities and limitations of modern AI face generation technology through hands-on experience.
Share your score on social media and challenge friends to beat your accuracy percentage.
The AI-generated faces in this challenge are created using advanced machine learning models, specifically Google's Imagen AI system. These systems are trained on millions of real photographs to learn the patterns and characteristics of human faces.
Modern AI face generators use a technology called generative adversarial networks (GANs) or diffusion models. These systems learn to create new images by studying vast datasets of real faces, understanding everything from facial proportions to skin texture, lighting, and even subtle imperfections that make faces look realistic.
The AI doesn't copy existing faces - instead, it synthesizes entirely new faces that have never existed in reality. Each generated face is unique, combining features in novel ways that the AI has learned are statistically plausible based on its training data.
In recent years, AI face generation has improved dramatically. Early AI faces from 2017-2018 had obvious artifacts like misaligned eyes, strange backgrounds, and unrealistic skin texture. Modern systems like Imagen can create faces that are virtually indistinguishable from real photographs, even to trained experts.
Studies have shown that humans can only correctly identify AI-generated faces about 50-60% of the time - barely better than random guessing. This presents both exciting opportunities for creative applications and concerning implications for misinformation and identity fraud.
Looking for more AI face-related content? Check out our other tools and resources:
Generate and download multiple AI faces at once. Perfect for design projects, mockups, and creative work.
Try Generator →Browse our entire collection of 67+ AI-generated faces. Download any face instantly for free.
View Gallery →Create professional AI headshots for LinkedIn, resumes, and business profiles with ToyRender.
Generate Headshots →Generate full-body AI models for fashion, e-commerce, and creative projects at This Model Not Exist.
Create Models →