I was flicking through a photo album at my grandma’s when I came across a picture of my mum as a child. I took a photo and sent it to her, but on my phone screen, it looked brighter and more vivid than the physical version in my hand.

Adding an Instagram filter is something I would now only do ironically. But is my phone increasing the contrast or making other tweaks without my knowledge? To find out, I downloaded an app with a “zero-processing” feature that claimed to take photos without any software alterations. When comparing the photos my camera takes automatically to the photos taken with this app, the results were shocking. The so-called “raw” photos that lack processing had subtle, muted colours, softer edges – a little grainy – while the processed photos were gorgeous and crisp like the inside of a marble. Why were they so different?

The answer to this, like everything else these days, is machine learning, used by virtually every major smartphone-maker to enhance the photos taken with their cameras. Professional photographers have been aware of it for years, as you can readily see on Reddit, YouTube or Facebook. On Aurora-Hunters UK, enthusiasts accuse one another of “cheating” by using phone cameras that “automatically brighten the picture”. But outside these niche circles, it’s rarely discussed. We communicate, build relationships, advertise ourselves through our pictures – and yet they are being heavily manipulated without our knowledge. I told my friend about this on a walk, showed them the two photos side by side, and on the train home they........

© The Guardian