AI channels get views for the same reason any video gets views: the idea and the packaging (the title and the thumbnail) are probably like 80% of the results on most videos, visuals that are boring as hell can still do really well, and once a video has views, the average viewer assumes it's the better video. What's changing is what YouTube keeps. In the January 2026 wave of terminations I went through, the channels getting wiped were fake AI news, fictional AI stories, low-effort spam and repeat-everything slop, while my own channels (which use some AI) were having record months.
I can't tell you how every viewer feels about AI. What I can tell you is what decides whether a video gets views, because I typically check it in the same order when I review a channel: the idea, the packaging, the audio quality, the intro, then the talking points. The thumbnail design and the packaging of the idea are probably like 80% of the results on most videos.
Notice what isn't on the list. Visual quality.
The audio is different. It has to be watchable, not so bad you can't stand listening to it.
Views are social proof. If 265,000 people watched one video, the average person on YouTube is going to think it's better than a 20,000 view video, even though the smaller one could have had better information.
| What I check, in order | What I'm looking for |
|---|---|
| 1. The idea | A topic this audience already clicks on, found from videos in the niche that already went viral |
| 2. The packaging | The title and the thumbnail, compared against the competitor that's winning |
| 3. Audio quality | Watchable. Not so bad you can't stand listening to it |
| 4. The intro | The first part of the video, right after the click |
| 5. The talking points | Is the value there, is the substance there |
| Not on the list: visuals | Boring visuals can still do really well |
Look at what's at the top of that list. Packaging, packaging, packaging. The 80/20 of going viral on YouTube is having good packaging, and idea is always alpha.
When I review a channel, I compare it to its competitors. What are they using in the thumbnail? Yellow arrows? A stroke around the arrow? A drop shadow? What's the font, the colors, the layout? You always want to copy what's working, down to the drop shadow.
Then I compare the video itself. It doesn't mean you need fancy editing or a thousand dollars a video. Is the value there? Is the substance there? Does it generate the same feeling as the competitor's video does? If yours is worse on any of those, why should it get the views?
A lot of the time the difference between a million view video and a 100 view video, on a new channel, is just a few small tweaks. Sometimes it's the packaging, sometimes it's the title.
Creativity is expensive. Whenever you're creative it requires a lot of effort to come up with things. People with an advanced skill set are usually a lot more boring with their strategies, and it's a good thing. If you're not getting views, it's ultimately a skill issue. (The full list of what I check is in why you're not getting views on YouTube.)
Yes and no. When I recorded this in January 2026, people on X were freaking out about a wave of terminations after YouTube's summer update (around July 15), the one that was going to take down a bunch of AI channels. I have a bunch of my own faceless channels. We even use some AI to produce the videos. We hadn't had any issues and we were making record months when I recorded this.
If anything, Google and YouTube love AI content. Veo, the AI video tool, is theirs. What I saw when I scrolled through a public list of terminated channels was a set of specific patterns:
| What was getting terminated | What we do instead |
|---|---|
| Fake AI politics and news (and fake space news) | My geopolitics channel is fine: real research, different perspectives, facts we take pride in getting right, nothing sensationalized |
| Fictional AI storytelling played off as real | Use AI to supplement what the video is about, not as the content |
| Raunchy kids content | It's disgusting. Good that those channels are gone |
| Pages built on other people's clips (fake Mr Ballen and Joe Rogan pages) | No compilation channels unless you have permission |
| Sexualized content, gambling | Obvious community guidelines problems |
| Low-effort spam (a ton of videos in a short period) | Quality first, then quantity |
| General AI slop (AI Pope videos, AI bodycam channels) | Videos that have value in them |
My general answer when I recorded this: if you just make good content you will be fine. The full channel-by-channel walk through is in YouTube AI channel terminations explained.
Because the cleanup is still going. YouTube is getting huge allegations of AI slop, so it's taking a stand, but there's definitely still weird stuff on the platform it's trying to clean up. When I recorded this in January 2026, I still saw a lot of AI storytelling channels that didn't have any issues, and a good amount of them on the terminated list. If you logged on to a kids channel, it was like one in three videos was AI storytelling.
I've seen other gurus online teach people to build those low-effort channels. I knew from day one it wasn't a good idea, and most of the ways I teach have nothing to do with what was on that list.
Can you do this stuff with AI? You can. You can use AI to do the scripts, the voiceovers, the editing and the thumbnails. I use AI pretty much every day. But if you're trying to make 30K a month on a channel, I don't see too many channels using AI for a lot of those jobs. You're always going to do better having humans make the content for right now, and using AI as a tool.
Some of my students are super nerdy AI tech guys, the smartest of the smartest, and they struggle to make $30,000 a month on their channels. To get there you need a team and you need people.
Can AI videos get monetized? They can. We have a lot of students making like 50 grand a month with AI videos. The way we make sure there are no issues: human voice actors, and sometimes the voice actor shows up on video for 5 seconds, just so it's obviously human. (More on that in can you monetize AI faceless YouTube channels.)
| Job | Who does it on my channels |
|---|---|
| Running it all | A channel manager I hire, like an operator |
| Script | A human script writer, using Claude or ChatGPT or whatever is the best AI that month |
| Voice | A human voice actor |
| Editing | A human video editor |
| Thumbnail | A thumbnail designer |
| The job posts to hire them | Upwork, which will throw up the job post for you in about 20 seconds |
Back in 2024 I said a good AI voice can sound better than a really bad voice actor with a crappy mic, as long as you don't use the generic baseline voice. When I was going through editor applications in April 2026 and a portfolio video played with an AI voiceover, my reaction was: I hate the AI voice.
Stop competing on effort and compete on the things that decide views. Find videos in your niche that already went viral and pick ideas the audience already clicks on. Copy the packaging that's winning to a tee (font, colors, layout), then put your own subject in it. Make the audio watchable. Compare your video to the competitor's one piece at a time until it gives the same feeling.
Then set standards and hold them. Most of the views on a lot of our videos come from television, so the videos have to be pretty quality. Raise the quality until it can't go up without spending a lot more money, then raise the quantity (three or five videos a week instead of one) by hiring more people for each role.
Some normal channels on the January 2026 list looked like they were flagged only because every thumbnail looked the same, as if they were reposting one video over and over. It's not the best thing for click-through rate, I know, but switch the thumbnail style every other video. Use the proven format, just not ten in a row that look exactly the same.
YouTube is competing with Netflix, TikTok, Instagram and X, and it wants real people and personalities, or at least real valuable content. If anything, I think YouTube is going back to a place where it's a good platform again. (Thumbnail detail is in the faceless YouTube thumbnail guide.)
Devon Canup says the idea and the packaging (title and thumbnail) are probably like 80% of the results on most videos, and videos with boring visuals can still do really well. Once a video has views, the view count works as social proof.
Not AI as such. When Devon went through the January 2026 terminated channel list, the patterns were fake AI news and politics, fictional AI storytelling, raunchy kids content, other people's clips, sexualized content, low-effort spam and general AI slop. His own channels use some AI and were having record months at the time.
Devon Canup says yes: he has a lot of students making like 50 grand a month with AI videos. They use human voice actors, and sometimes the voice actor appears on video for 5 seconds so the channel is obviously human.
Devon's answer is packaging. He compares a winning competitor's thumbnails in detail (arrows, drop shadows, font, colors, layout) and copies what works. On a new channel the gap between a million view video and a 100 view video is often a few small tweaks.
Devon Canup uses AI as a tool and humans for the content: script writers who use Claude or ChatGPT, human voice actors, editors and thumbnail designers. He says you can do the scripts, voiceovers, editing and thumbnails with AI, but he doesn't see many channels at 30K a month doing it that way.
Faceless Channel Academy is the coaching program I run for people building faceless YouTube channels: niche, formats, team, packaging, with an advisor and live calls. Book a free strategy call and we'll look at where you are and whether it's a fit.
Book a Free Strategy Call →Not ready to talk? Grab the free niche research template first.