Editor's take: As platforms embed AI deeper into everyday media consumption, few viewers may even notice the difference, and fewer still may reflect on how invisible alterations affect authenticity. The underlying concern is whether the push for polished presentation could gradually blur the line between authentic content and algorithmically mediated output, leaving audiences unsure of how much they can trust what they see.
YouTube has acknowledged that it has been using artificial intelligence to make video enhancements on its platform without first informing creators. The undisclosed changes, applied to a subset of YouTube Shorts, have left some content producers unsettled about how their work is being presented and the broader implications of invisible algorithmic mediation.
Rick Beato, a music educator and commentator with over five million subscribers, first suspected something unusual when his own appearance in a recent video looked subtly wrong. "I was like, 'man, my hair looks strange'," he told the BBC. Looking closer, Beato thought his skin and face appeared as if they had been digitally retouched, though he had not used any filter.
Another creator, guitarist Rhett Shull, reported similar distortions. After inspecting his own Shorts, he noticed oversharpened features and what he described as an artificial aesthetic. "If I wanted this terrible over-sharpening I would have done it myself," he said. Shull later posted a video about the issue that gained more than 500,000 views, warning that it risked undermining trust between creators and their audiences.
Their complaints echo earlier discussions on Reddit, where users shared evidence of unusual edits as far back as June. Among the issues highlighted: skin textures appearing unnaturally smooth, fabric folds looking exaggerated, and occasional warping of small details such as ears. These adjustments are subtle, usually requiring side-by-side comparisons to detect, but some creators say they are enough to raise suspicion about what else might be changing without their knowledge.
Responding to mounting speculation, YouTube's head of editorial and creator liaison, Rene Ritchie, stated in a post on X that the platform has been testing "traditional machine learning technology" to unblur, reduce noise and improve clarity in Shorts. Ritchie compared the process to enhancements already built into modern smartphones.
No GenAI, no upscaling. We're running an experiment on select YouTube Shorts that uses traditional machine learning technology to unblur, denoise, and improve clarity in videos during processing (similar to what a modern smartphone does when you record a video)
– YouTube Liaison (@YouTubeInsider) August 20, 2025
YouTube is always... https://t.co/vrojrRGwNw
The company has so far avoided saying whether creators will eventually be able to opt out of such changes. YouTube also stressed a distinction between "traditional machine learning" methods and generative AI, which produces new content altogether.
Hi! I'm a tech nerd and I try to be precise about the terminology I use
– YouTube Liaison (@YouTubeInsider) August 20, 2025
GenAI typically refers to technologies like transformers and large language models, which are relatively new
Upscaling typically refers to taking one resolution (like SD/480p) and making it look good at a...
Some researchers argue that YouTube's terminology downplays the scope of the changes. Samuel Wooley, professor and Dietrich chair of disinformation studies at the University of Pittsburgh, said the difference being drawn is less meaningful than it appears. "Machine learning is in fact a subfield of artificial intelligence," he said, adding that describing the process as merely a technical enhancement obscures the fact that AI is modifying the videos without explicit consent from their creators.
The question, Wooley argued, is not whether the system uses generative AI but whether the edits erode public trust in online content. "What happens if people know that companies are editing content from the top down, without even telling the content creators themselves?" he asked.
The episode comes at a time when consumer technology companies are increasingly incorporating AI into the media people create and consume. Google, YouTube's parent company, has promoted AI features in its Pixel smartphones, including Best Take, which can merge preferred facial expressions from multiple shots into one composite photograph. The newly released Pixel 10 can apply AI-assisted 100x zoom – an effect that exceeds the physical limits of its camera lens.
Other firms have experimented in similar ways. Samsung was accused in 2023 of artificially enhancing photos of the Moon taken by its Galaxy devices, later confirming the use of AI systems in the process. Netflix has faced criticism over AI-remastered versions of 1980s sitcoms that some viewers described as producing distorted and unsettling visuals.
YouTube confirms AI alterations to Shorts, raising concerns among creators
