![](/static/253f0d9b/assets/icons/icon-96x96.png)
![](https://lemmy.ml/pictrs/image/h1ChnLuBHr.png)
As someone else said, Pixel for Pixel analysis is probably too much compute time for them to bother. But they can do a quick checksum on the file, and they probably do.
Whether the image seriously affects your online fingerprint is mostly about whether a lot of users or only a few users use that exact profile picture.
If they few users have that exact profile picture, then it’s likely that they have a behavior tracking data set assigned to it, in case it’s valuable later.
It’s not that someone is sitting in a room correlating and judging your choice of picture. It’s just that every aspect of your web browsing that can be cheaply tracked and correlated is tracked and correlated.
An image that too many people use is likely also correlated, but won’t be heavily weighted in deciding that traffic is yours, because the error rate is too high.
That’s why I always set my profile image to “Mickey Mouse” while I listen to music by The Beatles. It makes me invisible. Also I just really like Mickey Mouse.
I don’t see what the big deal is. I store all kinds of sensitive information in plain text. SSNs, credit card numbers, birthdates and religious and political affiliation information.
The guy I bought it all from said it was okay, he stores it in plain text, too. (I’m joking, of course! Any information about you all that I’ve bought on the dark web, I’m storing responsibly.)