YouTube has begun quietly using artificial intelligence to enhance videos by some of its top creators—without notifying them or their audiences.
The practice was first noticed by two well-known American YouTubers popular among music enthusiasts: Rick Beato and Rhett Shull. Both run channels with millions of subscribers. Beato, a music educator and producer with more than 5 million subscribers, said he realized something was “off” in one of his recent videos.
“I thought I was imagining it, but my hair looked strange, and it almost looked like I was wearing makeup,” he said in a post.
Subtle changes, big questions
It turns out YouTube has been experimenting in recent months with AI-powered video enhancement, even altering YouTube Shorts without creator approval. The changes are subtle: sharper shirt folds, smoother or more highlighted skin, even slightly altered ears. Most viewers would not notice—but Beato and Shull said the edits made their videos feel unnatural.
Shull, a guitarist and creator, posted a video about the issue. “It looks like something AI-generated,” he said. “It bothers me because it could erode the trust I have with my audience.”
Complaints about odd changes to Shorts surfaced on social media as early as June, but only after Beato and Shull spoke out did YouTube confirm the experiment.
YouTube admits to ‘limited test’
Rene Ritchie, YouTube’s creator liaison, acknowledged in a post on X that the company was running “a small experiment on select Shorts, using traditional machine learning to clarify, reduce noise and improve overall video clarity—similar to what modern smartphones do when shooting video.”
That explanation drew further criticism. Samuel Woolley, a disinformation expert at the University of Pittsburgh, said the company’s wording was misleading. “Machine learning is a subset of artificial intelligence,” he said. “This is AI.”
The controversy highlights a wider trend in which more of what people see online is pre-processed by AI before reaching them. Smartphone makers like Samsung and Google have long used AI to “enhance” images. Samsung previously admitted to using AI to sharpen moon photos, while Google’s Pixel “Best Take” feature stitches together facial expressions from multiple shots to create a single “perfect” group picture.
“What’s happening here is that a company is altering creators’ content and distributing it to the public without their consent,” Woolley said. Unlike Photoshop filters or social media effects, he warned, YouTube’s AI edits add another hidden layer between audiences and the media they consume—raising concerns about authenticity.
Creators respond
While Woolley warned of eroding public trust, Beato remained more optimistic. “YouTube is always working on new tools and experimenting,” he said. “They’re an industry leader, and I have nothing bad to say about them. YouTube changed my life.”
Still, critics say even minor retouching without disclosure sets a troubling precedent. YouTube is home not only to entertainment, but also to news, education, and informational content—areas where accuracy and authenticity matter.
The quiet rollout suggests a future in which AI may increasingly reshape digital media before users even press play.