← All posts

Faceless Console v3.1: AI presenter channels, built inside YouTube’s rules

Faceless Jack · 7 October 2026 · 5 min read

Updated for v4.2: the presenter is never named and nothing public says AI. The only disclosure is YouTube’s “AI use” setting at upload, and Packaging still reminds you.

Version 3.1 of Faceless Console is out. There are no new skills — still twelve — but three changes run through the ones you have: AI presenter channels, a channel-art step at the end of the setup wizard, and a risk screen re-checked against YouTube’s current monetisation rules. Everything you already run works exactly as it did. Here is what changed and why.

Why this release exists

Demonetisation is the subject every faceless operator is talking about right now, and AI presenter channels — an AI avatar on camera, fronting the channel — are a big part of that conversation. YouTube’s monetisation policies are specific about both. Content must not be “mass-produced, generic, repetitive, or manipulative”. And channels that use AI-generated personas to deliver information on health, legal issues, finances or politics “will not be allowed to monetize”.

So v3.1 does two things. It lets you build an AI presenter channel inside those rules, and it makes the check that keeps you inside them dated and current.

AI presenter channels, always disclosed

The setup wizard now asks one more question at the end of the interview: should the channel speak in its own brand voice — the default, exactly as before — or through an AI presenter? Choose a presenter and the wizard designs one for the channel: a role (“presents the tech news desk”), a look, a voice, a personality and a sign-off, with a name from a shortlist made for the role and the audience.

The presenter is disclosed every time. A line in the channel description and in every video description says the presenter is AI-generated, the packaging skill reminds you to set YouTube’s “AI use” setting to Yes on upload, and by default the presenter says so in the intro. The script writer gives the presenter a voice and opinions, then runs a presenter pass on every script before it saves. A presenter never has:

  • an invented life story, credentials or years in a trade;
  • a fake family, friends or colleagues;
  • first-person memories presented as fact — “I tried it” becomes what the sources report;
  • first-person product endorsements;
  • a real person’s likeness, name or voice.

Presenters never give health, finance, legal or political advice, and a channel on those subjects can’t have a presenter at all. A disclosure line doesn’t change what YouTube’s rule is about, so the system blocks it and offers the same channel in brand voice instead.

Scripts are shaped for whichever avatar tool you use: one continuous read for most tools, or one sentence per line for clip-by-clip tools. Thumbnails can feature the presenter as the channel’s recurring face — at a desk or in a studio, never posed on location as if they had been there.

Bend an on-camera channel into a presenter channel

The Niche Bender now takes on-camera channels — a real person talking to camera — and AI avatar channels as sources, not just faceless ones. Before it builds anything, it sorts what the creator really offers. News, rankings, tips, warnings and explainers travel: someone else could read them out and the video would still deliver. First-hand experience — the tour, the room, the thing they tried — doesn’t travel truthfully, so it is left out or rebuilt as sourced reporting. And if people watch for the person themselves, the verdict is walk away.

Working the same lane as the on-camera channels, with an AI face, is allowed — but the face is never the difference on its own. The run has to find and name a real one: another country’s viewers, a different cadence, more depth, better sourcing. Same format, same audience, new face is a clone, and it is refused. Nothing of the creator is ever taken: not their name, look, catchphrase, set or likeness.

Banner, logo and keywords — the last step of setup

Channel art is now an optional final step of the setup wizard, and you can ask for it any time: say make my banner. It reads what the winning channels in your lane use, decides whether your art should be realistic, illustrated or graphic, and builds three directions. The banner is built to YouTube’s safe area, with previews of how it crops on TV, desktop, tablet and phone; the logo is tested at the 98-pixel size YouTube shows it. You also get your channel keywords, ready to paste and under 500 characters, and a setup document that opens with your channel description and keywords in copy-paste boxes.

The pictures can be made with your own Gemini or OpenAI key, from prompts you paste into a free image tool, or with no image tool at all. On a presenter channel the same step can make the presenter’s portrait. The Niche Bender’s logo and banner prompts are gone — this step replaces them.

The risk screen, dated and re-checked

The Niche Validator’s risk screen now states exactly which YouTube rules it screens against, quoted from YouTube’s own help pages and dated: checked 6 October 2026. It covers inauthentic content — templated, mass-produced, repetitive or manipulative videos — and AI personas on sensitive topics. If a run happens more than 90 days after that date, it re-reads the live pages first and tells you whether anything changed. The Niche Radar’s risk badge carries the same position, and the radar now flags lanes where NexLev lists channels that look like AI presenters, clearly labelled as unverified.

One thing worth knowing: YouTube’s AI label is for realistic content a viewer could take as real — an AI presenter, photoreal AI people or scenes. A normal AI voiceover over animation or stock footage doesn’t need it, and the tools say so. YouTube’s own page says disclosing AI content won’t limit a video’s audience or its eligibility to earn.

What it will not do

Promise monetisation. Nobody outside YouTube can, and YouTube makes the final call on every channel. What v3.1 does is build every channel around the rules YouTube publishes, designed to keep your risk low — and keep you on the right side of them.

If you already own Faceless Console, v3.1 is on your download page now. To update, delete your twelve old Faceless Console skills in Claude’s Settings → Customize → Skills first, then upload the twelve new ones and start a new chat — uploading over an existing skill doesn’t always replace it. Your channels aren’t touched. If you don’t own it yet, the price has not changed.

The Avatar Video Editor from this release is in the Editing bundle, with the Faceless Video Editor and the Shorts cutter. See the Editing bundle — $59 →

Already a licence holder?

v3.1 is on your download page now. Enter your purchase email and a fresh link comes back to you. Updating? Delete your old skills first, then upload the new ones — the steps come with your download.

Get v3.1