How to stop spam and scam comments on your YouTube channel
Summary
Every YouTube channel eventually meets the Crypto Bot: “I made $8,412 in 3 days thanks to…” No you didn’t, Chad. To stop spam and scam comments on my YouTube channel, I start with YouTube Studio’s comment filters and Held queue. I hold links and recurring scam phrases, remove deceptive comments, and report impersonators to YouTube. Then I add narrow custom rules for the patterns that keep coming back, and I check every match for legitimate comments before I reach for the hammer.
Key Takeaways
- Start with YouTube’s native controls. Check both Published and Held comments, including replies.
- A familiar avatar, repeated message, link or phone number needs context. None of them proves a scam on its own.
- Start every pattern rule on hold for review. Test it against warnings and useful links as well as suspicious examples.
- Report impersonation to YouTube and tell viewers which account and contact routes you use.
On this page
Which YouTube comment controls stop spam?
I start with YouTube Studio’s comment filters and Held queue before I build anything custom. In YouTube Studio, I choose Community and review both Published and Held. YouTube’s review guide explains these tabs. I inspect the replies under my own comments too, because a deceptive reply can look like part of my conversation. That is where the fake “giveaway admin” likes to lurk.
YouTube offers Basic, Strict, and Hold all moderation. Basic holds potentially inappropriate comments; Strict holds a broader range; Hold all requires your approval before any comment appears. Whichever level I pick, I review what the filters catch.
- For an existing video, open Content, select the video, then Show more. Adjust Comments and ratings and save. Upload defaults apply to new videos and posts; they do not update your older uploads.
- Open Settings, Community moderation, then Content controls. Turn on Link text to hold comments with hashtags and URLs. Add specific recurring phrases to Blocked words and save. Follow the current channel moderation settings instructions.
- Review the Held tab and approve legitimate comments. YouTube keeps held comments for up to 60 days. Approved users bypass word and link filters; your own comments and your moderators’ comments also bypass link blocking.
When a bot swarm shows up faster than I can review, I pause comments on that video. Existing comments stay up, and YouTube stops new ones until I turn comments back on. I run false-positive checks as I tighten the settings, so real viewers don’t catch a stray bonk.
How do you spot creator impersonation comments?
A creator impersonation comment copies your name or photo to trick your viewers, so compare the account’s handle and channel URL with your real account. Someone using my profile picture and a handle like “Andrew Weilbacher ✓ Official” is not me. A copied photo or similar display name does not establish identity. YouTube’s impersonation policy covers deceptive copying and comments that falsely suggest another person or entity wrote them.
I review every prize message that asks a viewer to pay a fee or move to a messaging app. I save the comment and account links somewhere private, then follow the reporting steps. I don’t open the linked destination to investigate; curiosity is how you get pwned.
I recommend a short notice from your own account that describes your real contact routes and giveaway process. For example, if it matches your practice: “I do not ask winners to pay a delivery fee in comments. Check the channel handle before replying.” Keep the notice accurate for your channel, and leave out the impersonator’s contact details. No free advertising for the scammer.
What should you do about repeated and copied comments?
Read repeated comments in context before you act: repeated praise can come from real viewers (hi, mom), while “sub4sub?” under every upload needs a different decision. YouTube’s spam policy covers high-volume repetitive or deceptive comments used to drive traffic or engagement, including sub-for-sub schemes.
I search YouTube Studio for a distinctive phrase and inspect the results. If I find the same unwanted solicitation across several videos, I remove those comments and consider hiding the account. Hide from channel applies across your channel. I check the account before acting, because similar text does not prove that two viewers are the same person.
A custom keyword rule can hold a recurring phrase. It matches the text of each comment; it does not count how often someone posted or detect duplicate campaigns. I keep a useful correction even when ten viewers point it out. If ten people noticed, I probably deserved it. For insults or criticism, see my separate guide to dealing with hate comments on YouTube.
How should you handle suspicious links?
Hold suspicious links for review with YouTube’s link-hold control before you build a long list of blocked domains. I judge a held comment from its wording and relevance without visiting the destination. A request for passwords, payment or off-platform prize collection gives me a reason to investigate and report; a source link that supports a correction may be useful.
The invented URL https://example.invalid/prize below is a non-working example. I never post real scam links in a public warning, and you shouldn’t either. A text rule that matches https?:// can find that marker, but it cannot tell whether a website is safe or catch every way someone might point to a destination.
I review the useful links my filters hold, such as a viewer’s source for a factual correction. If a rule keeps catching the discussion I want, I remove or narrow it.
Does a phone number in a comment mean a scam?
A phone number alone does not mean a scam; read the number together with the request. “Congrats, you were selected by Andrew! Message me on Telegram” is a scam wearing my face, and it needs scrutiny. A viewer discussing a public support number or quoting a scam warning may have a legitimate reason to include digits.
The example rule below looks for “WhatsApp” or “Telegram” followed by a number in a limited format. It holds matching text for a person to review. It cannot verify the account’s identity, and it misses other wording and formats. The number in the example is fictional; none of these examples comes from a real viewer.
My advice: write a rule for a recurring pattern you have seen on your own channel. Don’t block all numbers. Timestamps, prices and technical details belong in ordinary discussion, and “4:12” is not a crime.
How do you report scams and impersonation on YouTube?
- Save the comment link, the account’s channel URL, and a dated screenshot privately before clearing the comment. Redact personal details if you share evidence with someone helping you moderate.
- On YouTube, open the comment’s three-dot menu, select Report, choose the reason that fits, and submit. You can also use Report spam or abuse in YouTube Studio. YouTube reviews reports; reporting does not guarantee removal.
- For an impersonating account, open its channel description, select …more, then Report user and follow the prompts. Use the impersonation reporting route as well as reporting the deceptive comment.
- Remove the comment or hide the account when you need to protect your channel. Hiding and reporting have different scopes, and you can do both. Do not ask your audience to confront the account.
YouTube’s moderation guide separates removing a comment, reporting spam or abuse, and hiding a user. I would still use those native reporting controls even if another tool handles your routine holds.
How do you test custom moderation rules?
Test a custom moderation rule by starting it on hold for review, so you can inspect matches before you make an irreversible choice. These invented examples were tested against Moderaty’s source rule matcher. They show how text matching works, not scam-detection accuracy or results on a live channel.
| Rule type and pattern | Matching example | Review limit |
|---|---|---|
A recurring prize-fee phrasekeyword: pay a delivery feeAction: hold | You won a prize. Pay a delivery fee to claim it. | Also catches a viewer quoting that phrase to warn others. |
A recurring promotion phrasekeyword: sub4subAction: hold | Sub4sub? Visit my channel. | Also catches “Please stop posting sub4sub.” It does not count duplicates. |
A web-link markerregex: https?://Action: hold | Claim at https://example.invalid/prize | Also catches a useful source link. It cannot assess the destination’s safety. |
An app name followed by a numberregex: \b(?:whatsapp|telegram)[ :]+[+]?[0-9][0-9 ()-]{7,20}[0-9]\bAction: hold | Contact Telegram: +1 202 555 0147 to claim a prize. | Also catches a legitimate contact message. Other wording and number formats can pass. |
In Moderaty, keyword rules match a phrase without regard to case, and regex rules match a text pattern. A blocked-user rule takes the author’s channel ID, not a display name. The source matcher validates regex patterns and tries longer patterns first, so an overlapping rule can decide the action. I review my existing rules before adding one.
Protected handles skip rules and AI scoring, so I confirm an account before protecting it. For other retrieved comments, Moderaty applies matching rules before AI scoring. Its toxicity scoring does not establish that an account is genuine or that a link is safe.
How do you protect legitimate comments?
Protect legitimate comments by testing each pattern with comments you want to keep as well as comments you want to hold. I include a warning that quotes the scam phrase, a useful source URL, and an ordinary question. These invented cases show why a match needs a human, not just a regex with main-character energy.
- “They asked me to pay a delivery fee. Is that a scam?” matches the fee phrase while warning you about it.
- “The source is https://example.invalid/reference” matches the link marker without showing harmful intent.
- “The calculation at 4:12 should use 2026 prices” contains numbers but does not match the phone example.
- “Send a message to collect your prize” misses these rules. A non-match gives you no assurance that a comment is safe.
I check the Held queue after changing native filters. YouTube acknowledges that its filters can make mistakes. I approve legitimate comments and narrow phrases that keep catching useful discussion.
I check Moderaty’s action status as well as its review queue. A pending decision can stay public until the YouTube hold succeeds. Held and rejected comments can be restored from Moderaty’s audit log. Deletions are permanent, and you cannot lift an author ban in Moderaty.
How do you preview moderation on your channel?
Moderaty is a YouTube comment moderation tool, and its source implementation retrieves published top-level comments. It does not sweep YouTube’s Held queue. Access, quota, pagination and failures affect coverage, so I keep using YouTube Studio to inspect those areas. These limits come from the current retrieval source; this guide does not verify a hosted channel run.
If you want to take it for a spin, sign in to Moderaty, connect a channel you own or are authorized to manage, and open its Rules tab. Saving a rule adds it to active moderation, so I’d begin with a hold action. The channel page’s Dry run control previews a selected history window without changing comments on YouTube. It does not preview an unsaved rule.
You get one free moderation dry-run attempt per channel, used when it starts even if it fails. It spends no credits. I’d inspect the audit log and any partial-run status before relying on the results. The feedback preview is separate and covers one page of up to 100 comments. The channel-page source documents the moderation preview controls.
Channel connection requests Google’s youtube.force-ssl permission with offline access. This scope is broader than comment access alone, and I’d rather tell you that up front. Check the consent screen and Moderaty’s privacy information before granting access. You can revoke it through Google’s third-party access settings.
At the October 8, 2026 pricing check, hosted Moderaty costs US$5 per month for 100 AI classifications shared between moderation and enabled feedback digests. A digest classification can use an additional credit for the same comment. Rule and protected-handle moderation decisions use no credit. Self-hosting has separate infrastructure and OpenAI costs; the US$49 lifetime offer also uses your own OpenAI key and is limited to 1,000 lifetime places. Check current plans and pricing before purchasing, since they can change.
Frequently asked questions
Can I stop all spam comments?
No filter can guarantee that, and anyone who promises otherwise is selling something (probably crypto). I combine native holds, review and reporting, and I check for missed comments as well as legitimate ones a filter caught.
Does a phone number or Telegram mention prove a scam?
No. I check the account and the request. A rule match tells me which text matched; a person still needs to assess the comment.
Should I delete repeated comments?
Read their context first. Repeated corrections or questions may be useful. I apply my channel rules to repetitive promotion and report conduct that violates YouTube’s policy.
Does Moderaty replace YouTube’s spam review?
No, and I don’t want it to pretend otherwise. I keep reviewing replies and Held comments in YouTube Studio and report impersonators there. Moderaty’s source-confirmed retrieval and text rules have the limits described above.
For the wider workflow, my YouTube comment moderation guide for creators connects spam review with native settings, handling criticism and testing automation.