YouTube is rolling out a bold new experiment. The platform has started showing pop-up surveys to some users after they finish watching a video, asking one blunt question: “Does this feel like AI slop?”
1 YouTube is testing new pop-up surveys that ask viewers to rate whether a video they just watched felt AI-generated. The move signals the biggest crowdsourced push yet by any major platform to tackle low-quality AI content head on. And the internet already has a lot to say about it.
How YouTube’s New AI Slop Survey Works
2 Starting on March 17, YouTube watchers began seeing a new popup prompt during their videos. The question is direct and uses language that most internet users already know well. 3 On the mobile app, users are seeing a new pop-up when rating a video that asks, “Does this feel like AI slop?” or “How much does this video feel like low-quality AI?” 1 Users see five response options: “Not at all,” “Slightly,” “Moderately,” “Very much,” or “Extremely.” 7 The implementation varies across users, with some seeing a simple yes-or-no question instead of the five-point scale. That variation suggests YouTube is running multiple versions of the same test to see which format collects the most useful data.
YouTube has confirmed the test is real, but it has not said what it plans to do with the results. 1YouTube has not announced how widely it might roll out or if it will become a permanent addition.
YouTube AI slop pop-up survey viewer feedback test 2026
Why YouTube Is Crowdsourcing AI Detection
The timing is no accident. YouTube has been drowning in a rising flood of AI-generated videos for over a year now. And the numbers tell a troubling story.
8
Here is a quick snapshot of the AI slop problem on YouTube:
| Metric | Figure |
|---|---|
| Share of Shorts recommended to new users that qualify as AI slop | 21% |
| AI slop channels found among top trending channels | 278 |
| Combined views across those channels | 63 billion |
| Combined subscribers | 221 million |
| Estimated annual revenue | $117 million |
5 The platform has been drowning in AI-generated content for a while now, and its existing automated and human review systems haven’t kept pace. By turning to the viewers themselves, YouTube is essentially admitting that its algorithms alone cannot solve this problem.
YouTube CEO Called Out AI Slop as a 2026 Priority
YouTube’s leadership has been publicly vocal about the threat. 16YouTube CEO Neal Mohan said reducing “AI slop” and detecting deepfakes are priorities for the Google-owned video site in 2026.
In his annual letter published in January 2026, Mohan laid out the challenge plainly. 11“It’s becoming harder to detect what’s real and what’s AI-generated. This is particularly critical when it comes to deepfakes.”
11 He acknowledged the rise of AI slop and said YouTube remains committed to being “a place where people feel good spending their time.”
The platform has already taken action. 17In January 2026, YouTube wiped out 4.7 billion views in a single enforcement wave. Sixteen channels, with a combined 35 million subscribers, lost everything.
13 AI-generated content, due to YouTube’s updated rules, now expects human creativity in the mix. If your videos don’t show clear human judgment, intent, or value, the algorithm treats them like spam.
But enforcement is one thing. Knowing which videos actually feel robotic to real viewers is another. That is exactly where this new survey fits in.
The Big Debate: Quality Control or Free AI Training?
Not everyone trusts YouTube’s motives. Within hours of the survey going public, social media exploded with a very different theory.
4 User @birdabo lamented that the new YouTube prompt “sounds good until you realize they’re literally turning 2 billion users into unpaid AI trainers.” 4 “YouTube isn’t banning AI slop. They’re making you label it so they can train their next model to not look like slop,” claimed @TukiFromKL.
The concern is that Google, which owns YouTube, could feed the labeled data into its own video generation models like Veo. 5If Google feeds viewer-labeled “slop” data into its Veo video generation models, the end result could be AI that’s specifically trained to avoid the patterns humans flag as low quality.
The skepticism has historical roots too. 4In 2018, we found out that those reCAPTCHA bot detectors were using people’s answers and puzzle-solving skills to train AI. Google, YouTube’s owner, also owns reCAPTCHA.
“YouTube asking ‘does this feel like AI slop’ is not them protecting you. It’s them using you to train their next AI to make slop so good you’ll never be able to tell the difference.” – @barkmeta on X
YouTube has not addressed these claims directly. 3YouTube has not clarified how viewer ratings from the new pop-up will be weighted or used.
What This Means for Creators and Viewers
For creators, this is a wake-up call. 5If viewer sentiment directly influences how content is ranked or monetized, even legitimate creators using AI tools as part of their workflow could find themselves on the wrong end of a negative rating.
YouTube’s current policy does not ban AI content outright. 3YouTube currently doesn’t prohibit creators from using AI tools, nor does it require them to disclose AI-generated content, though they risk losing monetization if their content is flagged as low quality.
The key signals that put a channel at risk include:
- Mass-produced, template-driven uploads
- Robotic voiceovers with no original commentary
- Repetitive formats with no human perspective
- Visuals that don’t support real storytelling
12 YouTube is not against AI. It is only against confusion, spam and content that feels like it was assembled automatically.
Meanwhile, the regulatory landscape is shifting fast. 28New York Governor Kathy Hochul signed state legislative bills into law that will require explicit disclosure when AI-generated “synthetic performers” are used in advertisements, taking effect in June 2026. 28This law is consistent with similar disclosure requirements that will soon be imposed by the European Union AI Act in August 2026.
Platforms that can identify and label AI content at scale will have a serious advantage in meeting these new rules. YouTube’s viewer survey could be laying the groundwork for exactly that.
The AI slop crisis on YouTube is far from over, but one thing is now clear. The platform is betting that its billions of viewers can spot what its algorithms cannot. Whether this experiment leads to a cleaner feed or becomes another data pipeline for Google’s AI ambitions remains the million-dollar question. What is not in question is that viewers are tired of hollow, machine-made content flooding their screens. And for the first time, YouTube is handing them a way to say so.
What do you think about YouTube’s new AI slop survey? Is it a genuine effort to clean up the platform, or just another way for Google to collect free training data? Drop your thoughts in the comments below.
