YouTube's Secret AI: Reality-Bending Edits?

When Your Hair Looks "Strange"

Imagine you're a popular YouTuber, like Rick Beato, who boasts over five million subscribers and nearly 2,000 videos about music. You've just posted a new video, and something feels a bit off.

"I was like 'man, my hair looks strange'," he recalled. "And the closer I looked it almost seemed like I was wearing makeup."

He initially wondered if he was just imagining things. It turns out, he wasn't.

Over the last few months, YouTube has been secretly employing AI to subtly tweak videos. We're talking about things like shirt wrinkles appearing more defined, skin looking sharper in some areas and smoother in others, and even ears seeming to warp if you look closely. These changes are tiny, often imperceptible without a side-by-side comparison, but some YouTubers are finding the resulting "AI-generated" feel quite unsettling.

This isn't just a quirky YouTube experiment; it's part of a much broader trend where AI is pre-processing more and more of the reality we consume. The big question isn't whether we can spot these changes, but how this digital mediation might be eroding our connection to the actual world.

The AI-Generated Glitch

Rhett Shull, another well-known music YouTuber and a friend of Beato's, also noticed these strange artefacts in his own videos. He was quite vocal about it:

"The more I looked at it, the more upset I got," Shull admitted. "If I wanted this terrible over-sharpening I would have done it myself. But the bigger thing is it looks AI-generated. I think that deeply misrepresents me and what I do and my voice on the internet. It could potentially erode the trust I have with my audience in a small way. It just bothers me."

Shull's video on the topic quickly pulled in over half a million views, highlighting that this wasn't an isolated concern. Complaints had actually been bubbling up on social media since June, featuring close-ups of oddly distorted body parts and users questioning YouTube's intentions.

After months of speculation, YouTube finally confirmed they were indeed altering a limited number of videos on their short-form video platform, YouTube Shorts. Rene Ritchie, YouTube's head of editorial and creator liaison, shared in a post on X (formerly Twitter):

"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)."

He added that YouTube is constantly working on improving video quality and would consider feedback. However, they didn't respond to the BBC's questions about whether users would get to choose if their videos were AI-tweaked.

Consent, Control, and "Machine Learning"

Now, it's true that modern smartphones use AI for image and video enhancement. But Samuel Woolley, a leading expert in disinformation studies at the University of Pittsburgh, points out a crucial difference.

"You can make decisions about what you want your phone to do, and whether to turn on certain features," Woolley explained. "What we have here is a company manipulating content from leading users that is then being distributed to a public audience without the consent of the people who produce the videos."

Woolley also believes YouTube's choice of words, like "machine learning," might be a deliberate attempt to downplay the use of AI due to public concerns. While machine learning is a subfield of AI, he argues it's not a meaningful distinction in this context. Ritchie later tried to draw a line between "traditional machine-learning" and generative AI, which creates entirely new content. But for many, the fundamental issue of unconsented alteration remains.

What Does This Do to Our Relationship with Reality?

This whole situation highlights how AI increasingly inserts itself between us and the media we consume, often subtly. Jill Walker Rettberg, a professor at the Center for Digital Narrative at the University of Bergen, offers a great analogy:

"Footsteps in the sand are a great analogy," she says. "You know someone made those footprints. With an analogue camera, you know something was in front of the camera because the film was exposed to light. But with algorithms and AI, what does this do to our relationship with reality?"

It's a valid question. We've seen similar issues before, like the controversy in March 2025 surrounding an apparent AI remaster of '80s sitcoms The Cosby Show and A Different World on Netflix. Viewers noticed a "nightmarish mess" of distorted faces and backgrounds in what were supposed to be high-definition versions of shows originally shot on videotape. For more on how to identify such alterations, you can read about spotting AI video.

Even Google, YouTube's parent company, seems aware of these concerns. Their upcoming Pixel 10 phone will be the first to incorporate new industry-standard content credentials, essentially a digital watermark to identify AI-edited images. This aligns with broader discussions around AI ethics and responsible innovation.

However, Woolley worries that YouTube's unconsented AI edits risk blurring the lines of what we can trust online.

"This case with YouTube reveals the ways in which AI is increasingly a medium that defines our lives and realities," Woolley stated. "People are already distrustful of content that they encounter on social media. What happens if people know that companies are editing content from the top down, without even telling the content creators themselves?"

The Upside and the Uncanny

Not everyone is bothered, though. Rick Beato, for example, remains largely positive.

"You know, YouTube is constantly working on new tools and experimenting with stuff," Beato commented. "They're a best-in-class company, I've got nothing but good things to say. YouTube changed my life."

But this goes beyond just '80s reruns and YouTube videos. Remember when Samsung was found to be artificially enhancing Moon photos taken on its devices in 2023? Or Google Pixel's "Best Take" feature, which stitches together the best expressions from group photos to create a "perfect" moment that never actually happened? The upcoming Pixel 10 will even use generative AI to achieve 100x zoom, far beyond the camera's physical capabilities, making AI photo restoration a growing concern.

As these features become more commonplace, they force us to rethink what a photograph or a video truly represents. It's not an entirely new phenomenon; thirty years ago, people worried about the impact of Photoshop, and more recently, we've discussed the effects of airbrushing models and social media beauty filters. However, as Woolley suggests, AI seems to be putting these trends "on steroids." A recent study by the Pew Research Center highlights growing public concern about AI's impact on information integrity^[Pew Research Center - Public and AI].

The core issue here is transparency and consent. As AI continues to become an invisible hand shaping our digital experiences, how do we ensure we're still connected to an authentic reality, and who decides what that reality looks like?