What Is Ad Cloaking? A Beginner's Guide to Detecting Fraudulent Ads

What Is Ad Cloaking? A Beginner's Guide to Detecting Fraudulent Ads

Ad cloaking is a technique where advertisers show one version of a landing page to ad platform review bots and a completely different version to real users. The bot sees a clean, policy-compliant page. You see gambling, crypto, adult content, or scams. It's how fraudulent ads slip through the review process on Meta, Google, and TikTok — and it's a growing problem for anyone doing legitimate competitive ad research.

This post is the short, beginner-friendly explanation. If you want the full technical deep dive — with the story of how we discovered it inside AdLibrary, a breakdown of the detection stack, and what it means for ad platform trust — read our full investigation: Ad Cloaking on Meta.

The Simple Version

Imagine a bouncer at a club who lets everyone wearing a suit walk straight in. Someone figures out that if they wear a suit past the bouncer and then change into something else once inside, they can do whatever they want. That's cloaking, minus the suit.

In the ad world:

• The bouncer is the ad platform's review bot (Meta's automated policy checker, Google's Ads Review, etc.).
• The suit is a clean, policy-compliant landing page the scammer shows to the bot.
• "Inside the club" is what real users see after they click — usually something the platform would never allow.

The technical trick is deciding who's a bot and who's a real user. That's done with server-side logic that checks your IP address, user agent, cookies, referrer, and a few other signals. If you look like a bot, you get the clean page. If you look like a real human, you get the real offer.

Why Does Ad Cloaking Exist?

Because ad platforms pay scammers extraordinarily well, as long as the scammers don't get caught. A single cloaked ad campaign pushing crypto scams, fake supplements, or gambling to the wrong audience can generate tens of thousands of dollars in a day. The review system is the last line of defense. Cloaking bypasses it.

Legitimate advertisers — the ones reading this post — never use cloaking. But understanding how it works matters for three reasons:

1. Platform trust: If your category is flooded with cloaked competitors, ad platform algorithms start distrusting the entire vertical. Your CPMs go up. Your approval rates go down.
2. Competitive research: When you're using a tool like AdLibrary to study competitor ads, you need to be able to spot when an ad you're looking at is fraudulent — because copying a scammer's playbook is a very fast way to get your own ad account banned.
3. Consumer trust: Users who get burned by cloaked ads lose trust in the entire platform, which hurts every legitimate advertiser on it.

How Is Cloaking Different From Legitimate Personalization?

This is the nuance that trips a lot of people up. Legitimate ad platforms use personalization all the time — retargeting, A/B testing, dynamic product ads, geo-targeted landing pages. All of these show different content to different users. None of them are cloaking.

The difference is intent and disclosure. Personalization is transparent: the advertiser tells the platform what they're doing, and the variations are all compliant with platform policy. Cloaking is deceptive: the advertiser hides one version from the platform specifically so it can't be reviewed.

A rough test: if the ad platform would approve every version of your landing page if they saw it, you're personalizing. If they'd ban one of the versions, you're cloaking.

How to Spot a Cloaked Ad During Research

When you're doing competitor ad research in AdLibrary or any ad intelligence platform, a few red flags suggest an ad might be cloaked:

• Very short run time. Cloaked ads tend to get caught and killed within days. If an ad shows up once and disappears, it may have been flagged.
• Advertiser name doesn't match the landing page brand. A common cloaking pattern is using a shell account to run ads whose landing page is branded differently.
• Landing page domain is suspicious. New domains, domains registered in unusual jurisdictions, or domains with a history of abuse.
• Category mismatch. An ad that looks like a home goods ad but targets adult demographics is a classic cloaking signature.
• The ad previews fine in AdLibrary but the live landing page is dead or redirected. This is the most obvious tell.

If you think an ad might be cloaked, don't follow the landing page link from an unprotected browser. Use a sandboxed environment or an anti-fingerprinting browser.

What AdLibrary Does About It

AdLibrary is a legitimate ad intelligence platform, which means we surface ads as they appear on ad platforms — including ones that were cloaked at the time of indexing. We don't remove ads from the index for being low-quality or sketchy, because that would defeat the purpose of competitive research. But we do:

• Index ad platform review status changes, so you can see when an ad was removed.
• Show longevity signals prominently in ad timeline analysis, so you can filter out recently-killed ads.
• Provide AI ad enrichment that flags advertiser patterns — making it easier to spot shell accounts and category mismatches.

If you're using AdLibrary for serious competitive research, these signals let you separate legitimate long-running winners from flash-in-the-pan cloaked campaigns.

Further Reading

• The full story: Ad Cloaking on Meta — how we discovered it, how it works technically, and what it means
• Guide to competitor ad research
• Meta ad library guide: competitive intelligence
• AdLibrary features

TL;DR

Ad cloaking is when an advertiser shows a clean page to the platform's review bot and a completely different (usually fraudulent) page to real users. It's how scams slip past Meta and Google's ad review. Legitimate advertisers never use it — but understanding it is essential for competitive researchers, because copying a cloaker's playbook is a fast route to a banned ad account.

For the full technical breakdown and the story of how we ran into this while building AdLibrary, read the full investigation.