The Situation With YouTube Shadowbanning Is Terrible for Small Creators
Something happened to Kurzgesagt that should seriously concern anyone running a small or faceless YouTube channel.
Not because Kurzgesagt is a small creator.
Because they are the exact opposite.
They've been making videos on YouTube since 2013. They have more than 20 million subscribers. And their latest problem gives us a rare look at what can happen when an automated YouTube system gets something completely wrong.
And the really disturbing part is this:
Their video wasn't failing because viewers hated it.
It was barely being shown.
The problem may not always be that viewers rejected your video. Sometimes the bigger question is whether enough of the right viewers ever received it.
On July 31, Kurzgesagt publicly explained what happened.
They said some of their videos had started showing strange fluctuations in views.
Then their latest video about microscopic superpredators performed extremely badly.
But this wasn't a normal bad upload.
Kurzgesagt said it became their worst-performing upload since 2013.
That alone would be interesting.
But then they looked at the analytics.
People were clicking the video more often than usual.
People were watching it for longer.
And the viewers who watched it were responding positively.
Almost every important performance signal was apparently above average.
Yet the video still collapsed.
That's the part that matters.
Because normally, if a video gets poor results, you can investigate the obvious explanations.
- Maybe people didn't click.
- Maybe they clicked and left.
- Maybe the topic wasn't interesting.
- Maybe the thumbnail was weak.
- Maybe the opening was terrible.
But what do you do when the people who actually receive the video are responding well...
and the video still barely reaches anyone?
That's where this story gets much more uncomfortable.
What Happened to Kurzgesagt?
Kurzgesagt contacted YouTube.
According to the channel, YouTube confirmed that something was wrong.
And the explanation they received was that YouTube's automatic AI detection systems had incorrectly identified their very much human-made videos as AI-generated slop.
Kurzgesagt said that this caused the system to choke their channel.
Think about what that means.
A channel that has spent more than a decade building an unmistakable visual style...
A channel whose videos require enormous amounts of research, illustration, animation and production...
was apparently caught by an automated system designed to identify low-quality AI content.
And the result looked, from the creator's perspective, remarkably similar to what creators call a shadowban:
The content remained online.
The channel remained online.
The viewers who found it liked it.
But the reach collapsed.
That's why I think the Kurzgesagt situation is much more important than simply asking whether YouTube shadowbans creators.
The bigger question is:
What happens when YouTube's automated systems are wrong?
Because YouTube is dealing with an enormous amount of content.
And the platform is increasingly trying to identify and reduce low-quality, repetitive and mass-produced content.
The objective is understandable.
There is an enormous amount of low-effort AI-generated content flooding the internet.
YouTube has a legitimate reason to try to deal with it.
But automated detection has a fundamental weakness.
It can make false positives.
And Kurzgesagt appears to be an extraordinary example of exactly that problem.
The irony is almost absurd.
Their distinctive animation style is so recognizable that there are countless creators and AI systems influenced by it.
And now a system designed to detect AI-generated content apparently interpreted the original human creators as AI-generated.
That's not proof that every faceless channel is being shadowbanned.
But it proves something much more important:
An automated system can make a serious classification error and the consequence can be distribution.
Why Small Creators Should Pay Attention
And this is where small creators should pay attention.
Because Kurzgesagt had something most of us don't have.
Direct access to YouTube.
They could investigate the anomaly.
They could contact people at YouTube.
They could get an explanation.
They could get the problem looked at.
And according to Kurzgesagt, their contacts at YouTube were helpful and transparent once the issue was escalated. The channel said it hoped the bug was fixed for now.
Now imagine the same thing happening to a channel with 300 subscribers.
Or 30.
Or 16.
You upload a video.
You get almost no impressions.
Your CTR looks fine.
The people who do watch it stay for a reasonable amount of time.
You check everything you can think of.
Nothing obviously explains the distribution collapse.
So what do you do?
- You might conclude the topic was bad.
- You might change the thumbnail.
- You might rewrite your title.
- You might change your niche.
- You might assume your videos aren't good enough.
- You might even abandon the channel.
And you might never discover that an automated system incorrectly classified your content.
That's the scary part.
Not necessarily that YouTube is deliberately targeting small creators.
We don't have evidence for that.
The scary part is that a creator can experience a distribution problem without necessarily knowing what caused it.
And the smaller you are, the harder it is to prove that something unusual happened.
The Small-Creator Data Problem
A huge channel has historical data.
It has millions of previous impressions.
It has established audience behavior.
It has people monitoring analytics.
It may have direct platform contacts.
A tiny channel doesn't.
For a small creator, going from 5,000 impressions to 50 is an obvious anomaly.
But if you normally get 50 impressions...
and suddenly get 10...
what exactly are you supposed to conclude?
That's why the word shadowban becomes so dangerous.
It can describe a real-looking outcome without telling you the actual cause.
You could be suppressed.
You could have a weak topic.
You could have poor packaging.
You could be competing against stronger videos.
You could have a tiny audience.
You could have a recommendation mismatch.
Or an automated system could simply have made a mistake.
From the creator's dashboard, some of those situations can look surprisingly similar.
And I've experienced enough strange distribution patterns across my own channels that I don't think it's rational to simply dismiss the possibility.
But I also don't think it's rational to call every low-impression video a shadowban.
We need evidence.
What Kurzgesagt Actually Demonstrates
And Kurzgesagt gives us an important piece of that evidence.
Their case shows that strong viewer response does not necessarily guarantee normal distribution if another automated system interferes upstream.
That's the key lesson.
The recommendation system doesn't operate in a vacuum.
There are multiple systems interacting with your content.
And if one of those systems incorrectly classifies your video, the recommendation system may be working with a bad input.
You can then sit there staring at your CTR and retention wondering why the algorithm hates you.
But the problem may exist somewhere else.
And that creates a brutal situation for small creators.
You don't just need to make good videos.
You need the platform to correctly understand what your videos are.
Why Faceless Channels Deserve Extra Attention
And if you're running a faceless channel, that deserves even more attention.
Because faceless content often relies heavily on combinations of voice-over, stock footage, animation, graphics, AI-assisted tools and standardized production workflows.
Those characteristics can overlap with the surface-level patterns that automated systems may associate with mass-produced content.
That does not mean faceless content is automatically being suppressed.
It means the risk of being misunderstood deserves investigation.
So What Should Small Creators Actually Do?
First, stop treating low impressions as proof of a shadowban.
But also stop treating them as proof that your content is simply bad.
Look for patterns.
- Compare multiple uploads.
- Check impressions.
- Check traffic sources.
- Check CTR in context.
- Check retention.
- Look at whether the people who actually receive the video respond positively.
- Look for sudden changes from your normal baseline.
- Document everything.
Because one bad video proves almost nothing.
But a repeated pattern across multiple videos becomes evidence worth investigating.
And if something looks genuinely abnormal, don't immediately start changing your entire channel.
Preserve the data.
Compare the affected videos with your normal uploads.
And if you have access to YouTube support, ask specific questions rather than simply saying:
"Am I shadowbanned?"
Ask what changed.
Ask whether there are restrictions.
Ask whether the videos are being recommended normally.
Ask whether there is an automated classification affecting distribution.
You want an answer that can actually be tested.
The Bigger Problem
Because here's what Kurzgesagt has unintentionally demonstrated.
Even one of the biggest educational channels on YouTube can be misclassified by an automated system.
And if that can happen to them, we should at least acknowledge the possibility that smaller creators can experience similar false positives without ever knowing what happened.
That doesn't mean YouTube is secretly sitting there deciding which small creators deserve to fail.
It means something more uncomfortable.
The system can be wrong.
And when the system is wrong, the biggest creators have a much better chance of finding out.
The smallest creators may simply see their impressions disappear...
and assume they're the problem.
So What's the Hope?
I don't think the answer is to panic about AI detection.
And I don't think the answer is to stop using faceless formats.
The answer is transparency.
Better diagnostics.
Clearer explanations.
And a real path for creators to challenge serious automated mistakes.
Because if YouTube wants to use increasingly powerful automated systems to police billions of pieces of content, creators need some way to know when those systems get it wrong.
Kurzgesagt got lucky in one very important way.
They were big enough to get someone's attention.
Most small creators aren't.
And that's why I'm watching this situation very closely.
Because I don't know whether my own low-impression videos were ever affected by something like this.
You probably don't know about yours either.
But now we have a documented example showing that an automated YouTube system can apparently misclassify human-made content and produce a severe distribution problem.
So the next time your video gets almost no impressions, don't automatically say:
"YouTube shadowbanned me."
But don't automatically say:
"My video must be terrible."
Either.
Investigate.
Because sometimes the most important question isn't:
"Why didn't viewers like my video?"
It's:
"Did the right viewers ever get a chance to see it?"