There are several possible responses to the question, "Why are you messing with my face?".
You can ignore it. YouTube tried that, but as messed-with-faces piled up that option became less tenable. You could try flattery: "You know, I think that extra nose really makes you stand out from the crowd," but that's not really going to fly. Or you can come clean and admit that, yes we have messed with your face but not in the way you think and really you should be pleased. Which is what YouTube eventully did.
The issue arose when viewers of YouTube Shorts started noticing that some familiar faces looked a bit odd with eerily smooth skin and warped ears. Eventually the videos' creators noticed too.
"I was like 'man, my hair looks strange', said Rick Beatto, who makes videos about music and the music industry. "And the closer I looked it almost seemed like I was wearing makeup."
Another YouTuber complained of "terrible over-sharpening," saying his features looked AI-generated and that it "deeply misrepresents me and what I do."
At which point YouTube decided to play its trump card. We didn't use GenAI to change your features a spokesperson said, we used "traditional machine learning". Hah!
Well thank goodness for that. You were using good ol' trad ML to mess with our faces, not this newfangled AI stuff. That's right, said YouTube. Moreover, we're using traditional ML to "unblur, deionise and improve clarity" - and who doesn't want that?
If explaining which tools were used to do the face messing was meant to reassure the creators it didn't, as evidenced by the replies to YouTube's post on X. Because of course the question wasn't how did you mess with my face, but why did you do it without telling me or asking my permission. To this, answer came their none - at least not for a few days. YouTube now says it's working on an opt out for creators who don't want their videos 'improved'.
This is just the latest in a very long list of examples of social media companies asking for forgiveness not permission, and a relatively harmless one at that, particularly when compared Meta allowing chatbots to engage in sexualised chats with minors and xAI's Grok aping Hitler. But there are never any consequences so why would they trouble themselves to change, particularly with a White House intent on blocking any attempts at regulation of US companies.
Even Beatto, whose living depends on YouTube's capricious algorithms, was careful not to bite the hand that feeds, calling YouTube a "best-in class company". Mess with my face any time.