Human vision doesn’t “white balance” the way a camera does. The eye gathers light, but most of the color correction happens in the visual system: the retina, neurons, and especially the brain. This process is often called color constancy.

Your brain uses far more than raw sensor data. It analyzes context, compares different parts of a scene, and recognizes objects. A white wall looks white largely because your brain knows it should be white, even under warm or cool light. It can even adapt locally, effectively treating different parts of a scene differently.

A camera, by contrast, records RGB values from a sensor and applies a simpler global correction. Auto white balance estimates the illuminant from image data, but it usually lacks true scene understanding and object recognition at the level human vision uses. Camera sensors and display systems also do not respond like the human visual system, so additional color correction is needed just to produce a pleasing image.

So the short answer is: the brain, not just the eye, performs sophisticated color adaptation, and cameras can only approximate that with limited sensor data and algorithms.