A clone app copying FlickType's video and branding ran on the App Store for years, climbed the rankings on bulk-purchased fake ratings, and reportedly brought in close to two million dollars a year before Apple finally acted. That case is not an outlier. It is a useful starting point for anyone trying to understand why Apple's review process, which genuinely does screen every app before it goes live, still lets brand clones, counterfeits, and fleeceware apps through on a regular basis.
Apple's own description of App Store security sounds rigorous, and in several respects it is. Apps come from identified developers, submissions go through automated and human checks, and distribution is locked down with cryptographic protections against tampering. None of that is false. What it leaves out is what happens after an app clears review: how a clone can rank above the original it copied, how a subscription trap can run for years before enforcement catches up, and how an app can behave one way for a reviewer and a different way for everyone else. The cases below, from a cloned keyboard app to a knockoff emulator to malware that waited for a real SIM card before acting, are documented in detail and show where the review process's design creates openings rather than closes them. For security, fraud, and brand teams responsible for protecting a company's name on the App Store, understanding exactly where those openings sit matters more than trusting that Apple's review will catch an impersonator before a customer does.
How Apple's App Store review actually works
Apple's review process has three layers, and each one is built to catch a different kind of problem. The first is identity: every app comes from a developer who has agreed to Apple's guidelines and whose submissions are distributed with cryptographic guarantees against modification. That layer confirms who shipped the code and that nothing was altered in transit, but it does not confirm what the code actually does or whether the developer's name is itself misleading.
The second layer is human review of the listing itself, specifically the marketing text and screenshots, checked for accuracy. Apple designed this step to create a high barrier against the most common way malware gets distributed: pretending to be a popular app, or promising features the app doesn't actually have. That is a meaningful check, but it is a check on the listing page, not a full audit of runtime behavior or a confirmation that the app isn't impersonating someone else's brand identity.
The third layer is reactive rather than preventive: user reviews and reports after the app is already live. Apple has been direct about the limits of this system. widespread attacks against iOS, iPadOS and visionOS users impractical and uneconomical for financially driven attackers is the actual goal Apple states for its security measures, not elimination of every bad app. That framing matters for how security teams should treat the App Store: as a venue with meaningful but bounded protections, where some volume of impersonation and fraud is treated as an acceptable cost of doing business at scale, not a solved problem.
None of these layers claims to verify brand ownership. A developer can pass identity checks, write an accurate listing, and still be running an app that copies someone else's name, icon, or marketing without permission, because nothing in the published process checks a submission against another company's trademark or brand assets before approval.
Why fake ratings and reviews beat the review process
The clearest demonstration of that gap is KeyWatch, a watch keyboard app that copied FlickType's branding directly, including using video of the original app in its own marketing. According to developer Kosta Eleftheriou, who documented the case, KeyWatch launched as an empty shell with no real functionality, then used FlickType's growing popularity against FlickType itself by borrowing its look and its demo footage.
What made KeyWatch dangerous wasn't the copying alone. It was the fake rating and review activity bought in bulk from other scammers, which pushed KeyWatch's App Store ranking and review score above where an app with no real functionality should ever land. Eleftheriou reported that the app was making around two million dollars a year on the back of that manufactured trust signal. The same review score that's supposed to protect buyers from a bad download became the mechanism that convinced them to trust it.
Apple's response illustrates the limits of reactive enforcement well. Following Eleftheriou's findings, Apple removed KeyWatch and several other scam apps tied to the same campaign. But it left other apps from the same developer, a company called ReadIQ, untouched. One flagged clone coming down doesn't mean the operator behind it has lost their ability to publish the next one, and a security team watching only for a single app name will miss the pattern entirely.
Fleeceware apps take a different approach from KeyWatch. Instead of copying a specific brand's look, they rely on a free trial structure that quietly becomes an expensive subscription, counting on most users never noticing or never figuring out how to cancel. The scale of this category is larger than most brand teams expect. Fleeceware's economics work because churn is slow: a subscription that's merely annoying to cancel will still get paid for by users who mean to cancel it next week and never do. Brand, fraud, and security teams tracking financial harm to customers should treat this category as a baseline cost of the App Store's current review priorities, not a problem that's already been cleaned up.
In March 2021, a list of 133 fraudulent apps with an estimated 500 million combined downloads was published by Avast. More than a year later, 84 of those same apps were still live on the App Store, as documented by the VPN review platform VPNCheck. Those 84 apps alone were tied to over seven million downloads generating roughly eight and a half million dollars a month, and were collectively estimated to be tricking users out of more than one hundred million dollars a year. That gap of more than a year between Avast's flagged list and VPNCheck's follow-up check shows how long a fleeceware operation can keep collecting from users after being publicly identified, with no obvious change in Apple's enforcement pace in the meantime.
The mechanics are consistent across these apps: a free trial period with hidden costs attached, billing that continues even after a user tries to cancel or deletes the app, and sometimes a false claim that Apple specifically recommends the app. None of this requires copying another company's name or icon. It works because the App Store's review checks the listing text for accuracy at submission time, not the billing behavior a user encounters weeks later once the trial converts. That split matters for a brand's own monitoring strategy. Fleeceware doesn't need to copy a logo or a name to cause financial harm, which means a monitoring program built only to catch lookalike branding will walk right past it.
Apps that behave differently after approval
Some apps go further than hiding bad billing terms. They hide bad behavior entirely until they're confident a reviewer isn't the one running them. WIRED's reporting on a set of 18 malware apps found on the App Store documented exactly this mechanism: the apps waited to see a SIM card present on the device before contacting their command-and-control server and activating ad fraud behavior.
The logic is straightforward once you see it. A SIM card signals that the phone belongs to a real person using it day to day, not a test device or a reviewer's phone running through Apple's approval process. Apple's reviewers, by the nature of how review works, are more likely to be testing on exactly that kind of controlled device. The apps simply stayed quiet until that signal appeared.
A security expert quoted by WIRED explained why this specific tactic worked, pointing to a gap in what review prioritizes rather than a gap in review's rigor overall. Because ad fraud does not relate to actually malicious activity for the user, Apple likely does not put a high priority on policing it. Ad fraud doesn't harm the device owner directly the way malware or data theft does, so it sits lower on the list of things reviewers are actively hunting for, even though it still generates real revenue off real users. For comparison, security firm ESET separately found 42 adware apps on Google Play downloaded millions of times, which suggests this isn't an App Store-specific blind spot so much as a category of fraud that falls outside what either platform's review process is built to catch quickly.
The tactic has an implication beyond this one case. Any review process that relies on testing an app in a controlled environment is vulnerable to an app built to tell the difference between a controlled environment and a live one, and a present SIM card is only one of the signals available for making that distinction. For a security team evaluating any app on the App Store, including a clone of their own brand's app, the lesson is the same one these sandbox-evasion cases teach about malware generally: what an app does during review and what it does once it's installed on a real customer's phone aren't guaranteed to be the same thing.
Clone apps that copy another brand's name, icon, and developer identity
Counterfeit apps are the category closest to what most people picture when they hear "fake app." These imitate a real app's logo, screenshots, and artwork, and use a name that sounds almost identical to the original but with a letter or two changed. The listing description is often stuffed with the same keywords a user would type to search for the genuine app, which pulls the counterfeit into the same search results as the real thing.
Some counterfeits go one step further and copy the developer name itself, again with one or two letters altered, so that a user glancing at the listing sees a name that looks right at a quick read. This is a specific, repeatable tactic, not a rare edge case, and it's one of the more reliable signals that a listing is impersonating rather than competing.
A related and technically legal tactic is repackaging: taking open-source app code and modifying it, often by adding ads. Repackaging isn't illegal by itself, since open-source code is free to modify, but it misleads users about who actually built the app they're installing, and it's frequently paired with the same name and icon mimicry seen in direct counterfeits.
The iGBA case shows how all of this plays out against a real company. The developer of the long-running GBA4iOS emulator accused iGBA, a newly approved emulator, of being a knockoff. The timing made the frustration sharper: iGBA was approved around the same time the original developer's own competing app, distributed through Alt Store with Delta, had been sitting in review since March 5th despite Apple having just changed its guidelines to allow emulators in the first place. The knockoff added location tracking with no game-related purpose and loaded the free app with ads, neither of which had anything to do with emulating a Game Boy.
None of these tactics require hacking Apple's systems. They work within the review process rather than around it, which is why a brand can't treat "it passed Apple's review" as proof that a listing using its name, icon, or screenshots is legitimate. A user who recognizes a familiar icon and a name that reads correctly at a glance has no reason to dig further, and the fact that the listing cleared Apple's review in the first place only reinforces that trust.
Why Apple's removal reasons don't confirm brand impersonation
Apple did eventually remove iGBA, but the stated reason is worth reading carefully. Apple cited two specific guideline violations: copyright infringement and spam. Neither of those is an explicit finding that iGBA impersonated another brand or defrauded its users, even though the original developer's complaint was specifically about being cloned and overtaken in the approval queue.
That gap between removal reason and impersonation finding matters for any brand trying to use App Store takedowns as evidence of wrongdoing. Apple states that in dangerous cases involving fraud or malicious activity, it removes the app immediately and can notify users who downloaded it. That process describes a reaction to confirmed harm, triggered after the fact, not a proactive declaration that a specific company's brand was impersonated. A removed app tells you Apple found a guideline violation. It doesn't tell you Apple agreed with the impersonated brand's version of events, and it doesn't create a record a brand can point to later as confirmation of impersonation.
Building detection coverage for your own app's clones
Put the last several sections together and the pattern for a brand owner is consistent: Apple's review catches some things well, misses others by design, and removes apps reactively using reasons that don't necessarily name the harm a brand actually experienced. Waiting on Apple's own review cycle, or waiting for a developer to notice a clone and post about it publicly the way Eleftheriou and the GBA4iOS developer did, means a brand finds out about its own impersonation only after it's already been reported by someone else.
The signals worth watching for are specific and documented, not speculative. A developer name with one or two letters changed from the real one is a known counterfeiting tactic, not a coincidence. A listing with a suspiciously high volume of uniformly glowing reviews is itself worth a second look, since fake review generation is a documented tool fraudulent app creators use to build trust quickly. Both signals are things a brand can check for directly across app stores on an ongoing basis, rather than discovering only once a customer complains about being charged by an app they thought was official.
This is the same underlying problem Bolster AI addresses across other parts of a brand's digital risk protection program: a coordinated campaign against a brand's name, not a single rogue listing that disappears once reported. The same detect, investigate, connect, and act progression that applies to a lookalike domain or a fake social profile applies here, and it's the same logic behind Bolster AI's online brand protection software.
Detection means continuously watching app stores for the name, icon, and developer-identity patterns tied to a brand's actual apps, not scanning once after a tip comes in. Investigation means confirming whether a flagged listing is a genuine counterfeit, a legitimate competitor, or a repackaged open-source project before deciding how to respond. The same logic extends to domain monitoring and takedowns for lookalike domains and to social media brand protection for impersonating profiles, since the same operators often run all three at once.
Connecting a flagged app to the rest of a brand's external threat picture is what turns one clone into evidence of a broader campaign worth escalating. Bolster AI's analysts stay involved at that investigation stage specifically because distinguishing a real counterfeit from a false positive takes judgment an automated scan alone can't apply.
Deciding how seriously to treat a clone of your app
Not every clone warrants the same urgency, and treating all of them identically either overloads a security team or lets a dangerous one slip through alongside harmless ones. A few factors are worth weighing once a clone is confirmed.
Has the clone used purchased fake reviews to outrank the original app, the way KeyWatch did against FlickType? That's a sign the operator is actively trying to divert customers, not just copying code passively. Does the app request permissions or collect data beyond what its stated function requires, the way iGBA's location tracking had nothing to do with emulating a game system? That points to fraud risk beyond simple branding theft, since the operator may be harvesting data from users who believe they're using an official app.
Does the developer name or app name mimic the original closely enough to fool someone doing a quick search rather than a careful one? And has the clone already accumulated meaningful downloads or revenue before anyone caught it, the way the 84 fleeceware apps ran for over a year after being flagged? Each of these raises the urgency of escalation on its own. Together, they're the difference between a listing worth monitoring and one worth escalating immediately, rather than leaving the timeline up to Apple's own review cycle to sort out.
These factors don't just decide urgency within a single app store listing. A clone that's aggressively monetizing fake trust signals or harvesting data beyond its stated function is behaving like the fraud campaigns Bolster AI's phishing and scam protection and automated takedowns workflows are built to interrupt elsewhere, which is why an app clone belongs in the same escalation path as a lookalike domain rather than a separate queue a brand only checks occasionally.
References
"About App Store security - Apple Support (NZ)." support.apple.com, https://support.apple.com/en-nz/guide/security/secb8f887a15/web.
"Apple App Store is apparently still littered with malicious apps | TechRadar." techradar.com, https://www.techradar.com/news/apple-app-store-is-apparently-still-littered-with-malicious-apps.
"84 dangerous scam apps found on iPhone App Store stealing your money; check full list of apps | Tech News (HT Tech)." tech.hindustantimes.com, https://tech.hindustantimes.com/tech/news/84-dangerous-scam-apps-found-on-iphone-app-store-your-stealing-money-check-full-list-of-apps-71658128703429.html.
"Gameboy emulator is knockoff, says original developer." appleinsider.com, https://appleinsider.com/articles/24/04/14/app-stores-first-emulator-looks-like-its-an-ad-loaded-knockoff.
"How to Spot and Avoid Fake Apps." me-en.kaspersky.com, https://me-en.kaspersky.com/resource-center/preemptive-safety/identifying-and-avoiding-fake-apps.