Guide
YouTube demonetization rules: what limits ads in 2026, and what is fine
What decides whether a video earns full ads, limited ads or none: the 14 advertiser-friendly topics, the rules for the first seconds, titles and thumbnails, swearing, AI channels, and what to do when a video turns yellow. Every point is paraphrased from YouTube's own Help Center pages, linked at the end.
Checked against YouTube's English Help Center pages on 20 October 2026. YouTube changes these rules often; the change log is linked under Sources.
Three outcomes, and where to see them
YouTube's advertiser-friendly content guidelines give every video, Short and live stream one of three outcomes: it can earn ad revenue, it gets limited ad earnings, or it gets no ad earnings. Limited means brands that chose only safer content will not advertise on it, so it usually earns less. No ad earnings means it cannot run ads, though it may still suit other ways to earn on YouTube.
These are rules about ads, not about whether a video may stay up. Removals, age restrictions and strikes come from a separate rulebook, the Community Guidelines.
- In YouTube Studio, open Content and look at the Monetization column. The yellow icon means limited ads; filter by Limited to list every such video.
- Checking means no ads yet while systems review the video. It usually takes under 20 minutes and at most an hour.
- Since March 2025 a video first rated Limited or No ads can get an automatic extra review, sometimes by a person. Studio then says some checks are pending; most finish within hours, at most 24.
- The status usually settles within 24 hours, but it can still change later, for example after metadata edits or as viewers react to the video.
What YouTube checks
The guidelines cover the whole upload: the video, the thumbnail, the title, the description and the tags. Reviewers judge them together, and context matters most, followed by focus, tone, realism and how graphic something is.
- Titles and thumbnails are stricter than the video. Moderate swearing in a title or thumbnail limits ads, and the f-word there removes them. A sexual thumbnail removes ads on its own.
- The first seconds count. Graphic game violence in the first 7 seconds means no ads, and in the thumbnail it limits or removes them; from second 8 to 15 it can earn (September 2026).
- If the video informs or educates, say so in the title, thumbnail, description and tags. Reviewers use that context.
- Scheduled live streams are checked on their title, description, thumbnail and tags before they start.
- For channel reviews in the YouTube Partner Program, reviewers may look at the main theme, the most viewed and newest videos, the videos with the most watch time, the metadata and the About section.
The 14 topics at a glance
A short summary of every topic in the guidelines, from what usually earns full ads to what limits or removes them. The examples are YouTube's, shortened. Read the full page before you rely on an edge case.
| Topic | Usually earns full ads | Limits or removes ads |
|---|---|---|
| Inappropriate language | Bleeped or muted swearing, hell and damn, moderate swearing in the video, most swearing in music or stand-up, and since July 2025 strong swearing in the first 7 seconds. | Moderate swearing in the title or thumbnail, or in most sentences (limited). The f-word in the title or thumbnail, slurs and hateful language anywhere (no ads). |
| Violence | Police doing regular duty, mild violence, dramatized non-graphic violence, game violence after the first 7 seconds, non-graphic sports injuries. | Raw conflict footage, graphic descriptions of tragedies, dramatized severe injuries (limited). Ultra-graphic or glorified violence, graphic game violence in the first 7 seconds (no ads). Graphic game violence in the thumbnail limits or removes ads. |
| Adult content | Romance and kissing, non-graphic sex education, breastfeeding, dance that is not sexually graphic. | Sexualized titles or thumbnails, sexual pranks (limited). Nudity, sexual acts even if blurred, sexual thumbnails, teaching how to make explicit synthetic images (no ads). AI-generated visuals count too. |
| Shocking content | Mildly or moderately shocking material that is blurred or shown in an educational or documentary context. | Unblurred gore that still gives context (limited). Content made only to shock, including young-looking subjects in distress (no ads). |
| Harmful acts and unreliable content | Professional stunts in a controlled setting, fail videos without graphic injury, low-risk pranks, debunking misinformation, alcohol without glorifying it. | High-risk stunts, tobacco product reviews (limited). Dangerous challenges, unproven cures, vaccine or election misinformation, promoting tobacco or vaping (no ads). |
| Hateful and derogatory content | Discussing protected groups, or criticizing someone's views or actions without being hurtful. | Offensive material in news or education (limited). Hate or harassment toward people or groups (no ads). Calling it comedy is not enough. |
| Recreational drugs | Educational, comedic or music references and documentary portrayal; since September 2026 also drug use and sale in games and scripted content. | A focus on drug use or production without promoting it (limited). Promoting drugs or explaining how to buy, make or sell them (no ads). |
| Firearms | Standard guns in a safe setting, repair and maintenance, responsible airsoft or BB gun use. | Guns used outside a controlled setting (limited). Making or modifying guns, helping sell them, fully automatic weapons, minors using guns unsupervised (no ads). |
| Controversial issues | Prevention content, brief mentions, and since January 2026 non-graphic dramatized content on topics such as domestic abuse or self-harm. | Artistic, educational or documentary focus on these topics (limited). Graphic depictions, descriptive child abuse or eating-disorder content (no ads). |
| Sensitive events | Discussing a tragedy or loss of life without exploiting or dismissing it. | Profiting from or exploiting a disaster, attack, conflict or health emergency, even without graphic images (no ads). |
| Enabling dishonest behavior | Educational or comedic references, news reports. | Teaching hacking or unauthorized access, promoting cheating services such as essay writing (no ads). |
| Content for kids and families | Positive, low-risk content: healthy eating, sports, simple crafts. | Adult themes, any swearing or scary content aimed at kids (limited or none). Bullying, dangerous challenges, realistic guns (no ads). |
| Tobacco | Prevention messages. Documentaries and dramatized depictions are judged more leniently. | Promoting tobacco, vapes, e-cigarettes or related products (limited or none). |
| Incendiary and demeaning | Criticizing someone's opinions or actions without insulting them. | Shaming, bullying, singling someone out for abuse, claiming a tragedy was staged or its victims are actors (limited or none). |
Two rules catch people out. Strong swearing in the first 7 seconds no longer limits ads (July 2025), but swearing in the title or thumbnail still does. Graphic game violence is a separate rule: in the first 7 seconds it still means no ads.
Swearing: what changed and what did not
- July 2025: a video or Short with strong swearing in the first 7 seconds can earn full ad revenue. Before, it got limited or no ads. YouTube said it re-reviewed affected uploads and switched them to green without notice.
- Still limited: moderate swearing in the title or thumbnail, and swearing in most sentences of a video.
- Still no ads: the f-word in the title or thumbnail, and slurs or hateful language anywhere.
- Bleeped, muted or abbreviated swearing is treated as clean. Occasional swearing in music or background audio does not limit ads on its own.
- Separate from ads: the Community Guidelines' vulgar language policy can still age-restrict or remove videos with excessive profanity.
Is there a list of banned words?
No. YouTube does not publish a list of banned or demonetized words. In a September 2025 Creator Insider Q&A, a YouTube speaker said the company has resisted publishing one, partly because a list would read as an order never to use those words. The Help Center gives example words for moderate, strong and extreme profanity instead, and the examples differ between the US and UK pages.
Sensitive words such as killed, dead or suicide are not the problem on their own. Saying them once or twice does not limit ads, and there is no need to bleep them for that reason. What usually limits ads is a video that focuses on a topic such as suicide or sexual assault in a descriptive way. Putting the word in the title or thumbnail is allowed, but it signals that the whole video is about that topic.
Lists of demonetized words on other websites are guesses, not YouTube documents.
AI and automated channels
AI is not banned from monetization. YouTube's channel monetization policies name AI-assisted work among the allowed examples when the result shows your own idea and adds value. The risks are about how a channel is made, and these policies apply to the whole channel, not to one video.
| Policy | What it targets | What is allowed |
|---|---|---|
| Inauthentic content (renamed from repetitious content in July 2025) | Template-made or mass-produced videos with little variation, slideshows or scrolling text with little story, generic AI content without the creator's own insight. | A series with the same intro and outro where each video has its own story, focus or idea. |
| Reused content | Clips, compilations, readings or reactions without meaningful commentary or editing. It applies even with the original creator's permission. | Reviews, reactions with real commentary, a new voice-over, editing that makes the video clearly different from the source. |
| Unsatisfying or off-putting content | Emotionally manipulative formulas, shock for views, unrelated AI clips stitched together to shock, realistic fake disasters or celebrity deaths. | A cohesive story, your own spin on a trend, AI used to visualize a character and story you invented. |
| AI personas on sensitive topics | AI-generated personas presented as human experts giving health, legal, financial or political advice. | No exception is listed for these topics. |
Disclosure is a separate rule. When a video looks realistic and shows something that did not happen, makes a real person seem to say or do something, alters footage of a real place or event, or uses AI music as the main focus, set AI use to Yes in YouTube Studio when you upload. Animation, fantasy, beauty filters, captions, script help and cloning your own voice do not need it.
Since May 2026 the label for photorealistic AI content sits directly under the player (an overlay on Shorts), and YouTube can add a label itself when it detects significant AI use that was not disclosed. YouTube says a disclosure label on its own does not change reach or eligibility to earn, but repeatedly skipping a required disclosure can lead to penalties, including suspension from the Partner Program.
Recent changes to the rules
| Date | Change |
|---|---|
| October 2026 | Public-interest content, such as congressional proceedings and debates about ongoing government actions, can earn ads under the hateful content and harmful acts topics. |
| September 2026 | Drug use and sale in games and scripted content can earn ads. Graphic game violence from second 8 to 15 can earn; in the first 7 seconds it still cannot. |
| August 2026 | YouTube clarified how depictions of death in educational, documentary and news content are treated. |
| June 2026 | Using YouTube's own features or tools to create content meant to harm others became a serious violation that can restrict access to features. |
| March 2026 | Young-looking subjects, human or not, shown in distress with shock or gore cannot earn ads. |
| February 2026 | Firearm videos with 30-round magazines can earn ads; the limit was 29. |
| January 2026 | Non-graphic dramatized content about controversial issues can earn ads; child abuse or eating disorders as the focus still cannot. |
| July 2025 | Strong swearing in the first 7 seconds can earn full ads. Repetitious content was renamed inauthentic content to cover mass-produced videos. |
| March 2025 | Videos rated Limited or No ads can get an automatic additional review, sometimes by a person; decisions can take up to 24 hours. |
| December 2024 | Content that promotes or teaches making sexually explicit synthetic images cannot earn ads. |
Got a yellow icon? What to do
- 01Wait for the checks to finish. While Studio says some checks are pending, an extra review is already running; most finish within hours.
- 02Read the notice. In Studio, open Content and hover over the status in the Notices column; the Impact summary explains the decision.
- 03Request a human review only if you believe the video meets everything in the can earn ad revenue tier. Start it with the Action (Appeal) button on that notice.
- 04Expect up to 7 days. The decision arrives by email. Each video gets one appeal and the result is final.
- 05Do not delete and re-upload: it does not help. YouTube also says videos, titles and thumbnails edited in Studio to remove a problem are not eligible for another review, so fix problems before you publish.
- 06Give context. If the video informs or educates, make that clear in the title, thumbnail, description and tags; reviewers judge them together.
Strikes and channel suspensions are separate systems with their own deadlines. A Community Guidelines warning or strike can be appealed within 6 months, and a removal within a year.
Check before you upload
- Upload as private or unlisted first and wait for the checks before you publish. YouTube itself recommends this.
- Rate your video honestly in the self-certification questionnaire. After about 20 rated videos YouTube can judge your accuracy, and it relies more on the ratings of accurate creators.
- Check the title and thumbnail on their own, not only the video. They are judged more strictly.
- Check the script before you record. ViralDNA Policy Audit reads a script against the advertiser-friendly guidelines and the Community Guidelines, quotes every risky line and suggests a safer rewrite. No tool can promise full ads; YouTube makes the final call.