A few months ago, Danny Goodwin – editorial director at Search Engine Land and SMX – flagged a curious pattern on YouTube.
He wasn’t seeing content theft. He wasn’t even talking about clickbait.
What he’d noticed was this: creators uploading the same video multiple times, only with slightly different titles or thumbnails.
“Is this a strategy?” he asked. “Or is YouTube going to penalize it as duplicate content?”
It was a deceptively simple question.
But as I started digging, it became clear: what Goodwin had picked up on wasn’t just a YouTube hack or a content farm.
It was the visible sign of a much larger atmospheric disturbance.
In meteorological terms, it was a Fujiwhara effect – a phenomenon where two cyclonic storms begin interacting with one another, rotating around a shared axis, and sometimes merging into a single superstorm.
And on YouTube, those two storms were:
The explosive growth of YouTube Shorts.
The rapid rise of AI video generation.
What follows is the inside story of how those storms collided – from the algorithm tweak that triggered a surge of duplicate content to the July policy update that aimed to contain it – and what creators must now do to avoid being swept away in the next wave.
Storm 1: Hurricane Shorts
The first system began forming in September 2020 when YouTube launched Shorts in India, a tactical response to TikTok’s ban in the country.
It started small – vertical videos under 60 seconds – but quickly picked up steam.
Fast-forward to 2025, and Shorts isn’t just a feature. It’s the dominant format on YouTube.
Shorts now average over 70 billion daily views, according to YouTube for Press.
And Tubular Intelligence gives us a clearer snapshot of the dominance of the YouTube short-form format. In the last 90 days alone:
59.4% of YouTube uploads were Shorts.
This drove 87.7% of total views and 81.8% of engagements.
Shorts are mobile-friendly, easy to produce, and viral by nature.
They’re the perfect entry point for creators – but also the ideal format for something else: automation.
You don’t need a film crew to make Shorts. You don’t even need a camera.
You just need an AI video tool and a few prompts. Which leads us straight into the second storm.
Storm 2: Hurricane Sora
Enter Sora, OpenAI’s generative video model capable of turning text into photorealistic motion.
When demo clips of Sora dropped – showing cinematic drone footage of wild horses or a rainy Tokyo street scene created entirely from a prompt – it was like watching the sky darken before the deluge.
And Sora was just the most cinematic example. Along with it came a flood of creator-focused tools like:
InVideo AI: Optimized for speed and templates.
2Short AI: Designed specifically to crank out Shorts.
Quso.ai, Veed.io, and MakeShorts: Each has its own twist on automating short-form content.
These tools promised scale and speed – and they delivered.
The most popular video on InVideo’s YouTube channel is titled “I made 60 YouTube Shorts in 60 mins with just 2 AI tools.“
It has racked up over 2.1 million views and 56,000 engagement – double the platform’s average engagement rate.
The allure was clear: with AI, any creator could produce dozens – sometimes hundreds – of Shorts per week.
The only problem? They all started to look and sound the same.
The Fujiwhara Effect: March 31 to July 15
The moment these two storms collided was March 31.
That’s when YouTube made a subtle but significant change to how it counts views on Shorts.
Previously, a view wasn’t counted until a user had watched at least a few seconds.
But starting March 31, a view was logged the moment a Short began to play or replay.
This changed everything.
It meant that a looped 6-second video – built to autoplay – could rack up thousands of views even if no one watched it all the way through.
And when combined with AI’s ability to mass-produce video, the Fujiwhara effect occurred: mass duplication, instant views, and fast monetization.
Creators – and content farms – jumped at the opportunity.
July 15: The eye of the policy storm
By mid-summer, YouTube had seen enough.
On July 15, the YouTube Partner Program (YPP) made a “minor update” to one of its channel monetization policies.
To “clarify” that the “repetitious content” policy included content that is repetitive or mass-produced, the YPP renamed its policy “inauthentic content.”
This wasn’t a brand-new policy – but it was the clearest signal yet that YouTube recognized the Fujiwhara effect had caused the two storms to curve, merge, or change intensity.
Sarah from TeamYouTube posted a detailed response to creator questions about YPP policies in the YouTube Help Center. She clarified:
This isn’t new. Content that’s spammy, repetitive, or low-effort has always been ineligible for monetization.
AI is allowed. Creators can use AI tools – if they add originality, voice, or creativity.
“Mass-produced” means risk. Slight variations on the same video won’t cut it.
She also gave a couple of examples of inauthentic content:
Videos that only include readings of content you didn’t create yourself, such as website articles or news updates.
Music tracks that have only been altered in pitch or speed, but are otherwise unchanged from the original.
Repetitive videos with little to no educational content, commentary, or storytelling, and minimal differences between them.
Large volumes of videos created using the same format or template.
Slideshows or scrolling text with little or no narration, commentary, or informative content.
The message was clear: If you’re gaming the system with template-based duplication, you’re no longer eligible for monetization.
Dig deeper: How to make engaging long-form YouTube videos
Get the newsletter search marketers rely on.
See terms.
The creator backlash (and breakdown)
The policy hit fast – and hard.
Faceless AI channels began sounding the alarm.
On July 13, InVideo for Content Creators uploaded a video titled, “YouTube’s New Policy Just Killed Faceless AI Channels?”
In its first three days, it got 25,100 views and 1,053 engagements.
Why?
Because many of these channels had gone all-in on automation:
AI voiceovers.
Stock footage.
Keyword-optimized text.
Looped visuals.
And now they were learning that content like that – even if copyright-safe – could be labeled “inauthentic” and demonetized.
YouTube’s editorial and creator liaison Renee Ritchie stepped in to clarify the clarification.
This wasn’t about killing AI channels, he said.
It was about promoting content that had real value for viewers – entertainment, insight, education, or personality.
What still works: Value-driven reuse and AI creativity
Here’s the good news: YouTube isn’t banning AI or reused content.
What matters is what you add to it.
Shorts that remix content or reuse clips are still eligible if you’re adding substantial originality. That could be:
Commentary and critique (e.g., CinemaSins or LegalEagle).
Narrative overlays (e.g., storytelling with clips and custom voiceovers).
Educational framing (e.g., sports analysis or science explainers).
Humor and editing (e.g., meme-style remixing).
For example, CinemaSins uploaded a Short in April 2024 called “But if you can make is past the burning sea… strangeworld” – a 14-second riff on Strange World.
It racked up 3 million views and 310,000 engagements, with an engagement rate 6.2 times higher than the platform’s average.
That’s what real added value looks like. And YouTube still rewards it.
Your navigation guide: How to survive (and thrive)
If you’re a creator caught in this swirling environment of policies and platforms, here’s how to chart your course:
Use AI as a tool, not a crutch
AI should assist your creativity, not replace it. Use it for scripts, voiceovers, or visuals – but infuse every video with your unique voice or brand.
Avoid templated spam
Reusing the same structure, visuals, or voice with slight tweaks? That’s a red flag. If all your Shorts feel interchangeable, you’re vulnerable to demonetization.
Add substance
Comment, analyze, tell a story. Even 60-second content can have depth. The more engaging and informative your videos are, the safer – and more successful – you’ll be.
Diversify formats
Don’t rely only on Shorts. Pair them with long-form content or live streams. Build a multi-format brand that resists single-policy risks.
Stay informed
YouTube updates its policies regularly. Follow TeamYouTube. Join creator forums. And yes – keep reading Search Engine Land.
The core incentive: Views vs. watch time
Let’s not forget what really triggered all this: the change in how views are counted.
In 2012, YouTube prioritized watch time over clicks, a brilliant move that forced creators to stop chasing thumbnails and start making engaging content.
But the March 2025 shift – counting a view the instant a Short begins – feels like a step back. It incentivizes quick starts, not quality.
And until that changes, the temptation to flood the platform with “AI slop” will persist.
YouTube may be playing Whac-A-Mole with bad actors now – but the root cause lies deeper: the view-count model itself.
Final forecast: Creators, stay human
The YouTube Fujiwhara effect isn’t over.
The storms of Shorts and AI are still circling. And for creators, the challenge is clear: rise above automation by leaning into authenticity.
Those who focus on substance, storytelling, and value will endure. Those who chase speed, shortcuts, and spam will be swept away.
In a world where algorithms shift and tools evolve by the week, the most resilient strategy remains timeless:
Be human.
Be original.
Be worth watching.
Dig deeper: 3 ways to make sense of YouTube’s messy attribution