Article contents:
- Introduction: “My hair looks weird”
- YouTuber discovery: Hidden and unwanted edits
- YouTube confirmation: An “experiment” to improve quality
- The consent issue: Manipulating content without permission
- A broader trend: When AI mediates our reality
- Eroding trust: Is what we see real?
- Conclusion: A fine line between enhancement and falsification
Introduction: “My hair looks weird”
Rick Beato, who runs a YouTube channel with more than five million subscribers, noticed something strange about his face in one of his recent videos. He says: «I was telling myself, “Man, my hair looks weird.” And when I looked closer, it looked like I was wearing makeup». At first, he thought he was imagining things. But he wasn’t. It turned out that YouTube was secretly using artificial intelligence to edit users’ videos without their knowledge or permission, in a move that raises deep questions about our relationship with digital reality.
YouTuber discovery: Hidden and unwanted edits
Beato wasn’t alone. His friend Rhett Shull, another popular music YouTuber, began examining his own posts and noticed the same bizarre edits: shirt wrinkles appearing more defined, skin looking sharper in some places and smoother in others, and even ears occasionally seeming distorted. These changes are subtle, barely noticeable without a side-by-side comparison. But for some YouTubers, they give their content an unsettling and unwanted “AI-generated” feel. Shull says: «The more I looked at it, the angrier I got… I think this misrepresents me and potentially erodes the trust I have with my audience».
YouTube confirmation: An “experiment” to improve quality
After months of rumors in comment sections, the company finally confirmed that it is modifying a limited number of videos on YouTube Shorts. René Ritchie, head of creator liaison at YouTube, said in a post on X: «We’re running an experiment on select YouTube Shorts that uses traditional machine learning to remove blur, noise, and improve clarity». He added that YouTube is always working to provide the best possible video quality and will take feedback from creators and viewers into account.
The consent issue: Manipulating content without permission
Although modern smartphones come with built-in AI features to enhance photo and video quality, the situation here is entirely different, according to Samuel Woolley, a disinformation expert at the University of Pittsburgh. Woolley says: «You can make decisions about what you want your phone to do… What we have here is a company manipulating content from leading users that is then distributed to a public audience without the consent of the people making the videos». Not giving users the option to opt in or opt out of these edits is the core of the problem.
A broader trend: When AI mediates our reality
This incident is not isolated, but rather part of a broader trend where a growing slice of our reality is processed by AI before it reaches us. In 2023, it was discovered that Samsung was artificially enhancing moon photos. Google Pixel phones provide a feature that uses AI to “fix” smiles in group photos, creating a nice image of a moment that never actually happened in the real world. As these features spread, the question becomes not just whether you can tell the difference, but whether this technology is eroding our bonds with the world around us.
Eroding trust: Is what we see real?
Thirty years ago, there was a similar concern about the chaos Photoshop would bring to society. Today, AI puts these trends on “steroids”. Woolley says that YouTube’s use of machine learning to edit videos without user knowledge risks blurring the lines of what people can trust online. «People already don’t trust the content they encounter on social media. What happens if people learn that companies are editing content from the top down, without even telling the creators themselves?».
Conclusion: A fine line between enhancement and falsification
The YouTube case reveals the growing ethical challenge we face in the age of AI. Where do we draw the line between quality enhancement and reality manipulation? Imposing edits without consent not only harms the artistic integrity of creators, but also undermines the foundation of trust upon which these platforms were built. As AI continues to mediate our world, transparency and choice are more important than ever to preserve our shared connection to real life.
Frequently asked questions
Q: What edits did YouTube make to the videos?
A: YouTube used AI to make subtle edits such as increasing image sharpness, reducing noise, and improving detail clarity in YouTube Shorts, giving some videos an unnatural appearance.
Q: Were content creators notified of these changes?
A: No, they were not notified or asked for their consent, which angered many of them who felt their content had been tampered with.
Q: What is the difference between this and what smartphones do?
A: On smartphones, the user typically has the option to turn enhancement features on or off. In YouTube’s case, the edits were imposed by the platform on the content after it was uploaded.
