08 = Newsletter Individual #8

Issue #8

Jun 25, 2026

8 min read

YouTube Is Removing AI Content That Isn’t Disclosed. Are You Protected?

My plain-English breakdown of what triggers YouTube’s disclosure requirement, what does not, and how to build compliance into your workflow before the platform builds it for you.

Subscribe for Free

Share this Issue

What you’ll get in 3 minutes

TLDR

In this issue

8 min read

Big Idea

Understand

Check

Disclose

In January 2026, YouTube carried out its largest single enforcement action against undisclosed AI content. Channels with a combined 4.7 billion lifetime views and 35 million subscribers were removed in one wave. The policy itself is not new. YouTube introduced it in November 2023, expanded it significantly in May 2025, and reached full enforcement strength entering 2026. What changed is not the rule. What changed is that YouTube stopped warning and started removing. The creators who lost channels were not necessarily using AI maliciously. Most were using it without understanding where the line was.

“YouTube is not anti-AI. It is anti-deception. The line is not what tool you used. It is whether a viewer could reasonably mistake what was generated for what was real.”

The rule is simpler than most creators think, and it is misunderstood in both directions. Some creators are disclosing content that does not need to be disclosed. Others are skipping disclosure on content that clearly requires it. The enforcement pattern is consistent: content that makes a real person appear to say or do something they did not, or that presents a fabricated event as real footage, is the target. AI voiceovers, AI scripts, AI thumbnails, and AI-generated illustrations that do not make that claim are not. Knowing the difference is a five-minute read that could save a channel years of work.

The 4-Step Workflow

Step 1

Know What Requires Disclosure

The rule targets deceptive realism, not AI use in general.

YouTube requires disclosure when content is meaningfully altered or synthetically generated in a way that could be mistaken for real. Specifically: AI video footage of a real person saying something they did not say. AI-generated scenes presented as real events or real places. Cloned voices of specific real people delivering statements they never made. Realistic deepfakes of any identifiable individual. If your content could make a reasonable viewer believe something real happened that did not, disclose it. That is the full trigger.

All AI use → Deceptive realism only

Step 2

Know What Does Not Require Disclosure

YouTube explicitly exempts most of what creators use AI for every day.

YouTube’s 2026 guidance is explicit on exemptions. AI-generated scripts, outlines, titles, and descriptions do not require disclosure. AI voiceover that does not clone a specific real person does not require disclosure. AI thumbnails, illustrations, and clearly stylised or animated visuals do not require disclosure. Colour correction, background blur, noise reduction, upscaling, and other production assistance do not require disclosure. Cloning your own voice for voiceover or dubbing does not require disclosure. Most day-to-day AI use in a creator workflow is in the clear.

Panic disclosure → Accurate exemption understanding

Step 3

Build Disclosure Into the Upload, Not the Description

The label lives in YouTube Studio, not in your video text.

Disclosure happens in YouTube Studio during the upload process, not in the video description or verbal disclaimer. Open the upload details, find the Altered or Synthetic Content section, and select Yes when the content contains realistic AI-generated or meaningfully altered material. YouTube then adds a label visible to viewers below the video. This label does not reduce reach or monetisation eligibility. YouTube has confirmed that properly disclosed AI content is not penalised in recommendations. Over-disclosure carries no penalty. Under-disclosure carries four escalating consequences, ending in permanent YPP suspension.

Description disclaimer → Studio upload label

Step 4

Add One Question to Every Pre-Upload Review

A 30-second check that replaces a four-level enforcement process.

Before every upload, ask one question: could a viewer reasonably believe something real happened here that was actually generated or materially changed? If yes, disclose. If no, proceed. Add this question to the end of your Descript edit review as a checklist item before export. It takes 30 seconds. The enforcement ladder it replaces starts with forced labeling, moves through content removal, and ends with permanent loss of YPP access on every video in your channel, not just the flagged one.

Post-upload risk → Pre-upload 30-second check

Tool of the Week

featured

Descript

Transcript-based video editing with built-in review and export workflow

The feature I use for compliance review is Descript’s scene-by-scene transcript view. Before I export anything, I read through the transcript and flag every section where AI generated or meaningfully altered the content. This is faster than scrubbing the timeline because the transcript makes the content decisions visible as text rather than frames. Descript’s Overdub feature also handles voice replacement cleanly for sections where a retake is faster than a full regeneration. For creators who edit by transcript, compliance review fits naturally into the same session as the edit.

Try it yourself

Prompt Recipe

Copy & Use

You are a YouTube compliance reviewer for an AI video creator.

Context: I am preparing to upload a video to YouTube. The video was produced using the following tools and methods: [LIST TOOLS: e.g. Veo 3.1 for B-roll generation / ElevenLabs for voiceover / Descript for transcript editing / Runway Gen-4.5 for character scene / AI script drafted in Claude].

Task: Review each production element listed and classify it using YouTube’s 2026 disclosure rules:

REQUIRES DISCLOSURE:

— Realistic AI-generated footage of real people saying things they did not say

— Fabricated scenes presented as real events or real places

— Voice clones of specific identifiable real people

DOES NOT REQUIRE DISCLOSURE:

— AI scripts, outlines, titles, captions, descriptions

— Generic AI voiceover not cloned from a specific real person

— AI thumbnails, illustrations, stylized or animated visuals

— Production assistance: colour correction, noise reduction, upscaling

— Cloning your own voice for voiceover or dubbing

Format: Return a simple table. Column 1: Production element. Column 2: Disclosure required? Yes / No. Column 3: Reason in one sentence.

Final line: State whether the upload requires the Altered or Synthetic Content label in YouTube Studio.

Constraints: Apply the rule strictly. Do not recommend disclosure for exempt content. Do not skip disclosure for content that meets the trigger. Flag any borderline cases with a note.

Pro Tips:

Mini Case Study

Real Results

AI content uploaded without disclosure review. Channel flagged, forced label applied by YouTube. Policy warning issued.

After: Pre-Upload Compliance Workflow

One-question pre-upload review built into Descript export session. No flags, no forced labels, no warnings since implementation.

Download

FREE

AI Content Disclosure Checklist for Creators

Step-by-step checklist to review every upload for YouTube AI disclosure compliance.

What's included:

Free for newsletter subscribers. By downloading, you agree to our terms of use.

Vani Aggarwal

Filmmaker & Head of Content Strategy exploring the edge where Al meets
story.

Get Al & filmmaking hacks weekly.

Join 12,000+ creators getting actionable Al workflows, tool reviews,
and creative techniques delivered every Wednesday.

No spam, ever. Unsubscribe with one click. Read our privacy policy.

Share & Discuss

Spread the word

Filmmaker & Head of Content Strategy
exploring the edge where Al meets
story.

Start a conversation

Filmmaker & Head of Content Strategy
exploring the edge where Al meets
story.

Reply by email

Related Issues

Explore more AI video workflows and creative techniques from our archive

Issue #5

The A24 Effect: What the Google DeepMind Deal Means for You

What a $75 million studio deal signals about creative control and where independent filmmaking is heading.

Issue #7

You're Prompting AI Video Wrong. Here's the Fix.

Veo, Runway, and Kling each respond to different prompt structures. Here is what each model is listening for.

Issue #9

How to Use AI B-Roll to Fill the Gaps in Footage You Already Shot

Runway's Aleph model can reframe, relight, and re-weather a scene from text. The practical workflow for finishing a video without going back on location.

Issue #17

Designing with intention

Learn how to bring purpose into every creative process you start.

Ask Your Questions

Frequently Asked Questions

How to enroll for a Course?

All courses are available at vaniaggarwal.com/courses. Browse the full catalogue and enroll directly on each course page.

Can I get the recordings of my previous lectures?

The newsletter is a standalone publication, not a course. For recorded lecture access, visit your enrolled course dashboard on the courses page.

Who would be the instructor for enrolled course?

All courses are taught by Vani Aggarwal, filmmaker and AI strategy consultant with 16+ years directing across six countries and hands-on testing across 50+ AI production tools.

What kind of placement support will be given post completion of program?

Course graduates receive access to the community, resource library, and direct pathways to Vani’s consulting and coaching programmes for continued support.

GET IN TOUCh

Let’s Create Something –
Extraordinary

Whether you’re a creator building a compliant AI workflow, a production team navigating platform policy, or a brand looking for a sharper content strategy that holds up under scrutiny, this is where the conversation starts.