Skip to main content
0%

Dangerous Apps: Check Risks in 5 Categories

12 min readUpdated

Check what data an app receives, why it requests permissions and who it shares information with. Below are five categories: messengers and social networks, maps and transport, email, fitness and health, utilities and entertainment. Belonging to a category does not by itself make an app malicious.

Sources were checked on October 26–27, 2026. This article reviews documents and published cases; it does not test apps on your phone. Further reading: how data is used to manipulate users and why deleting an account does not fully erase your data.

Risks depend on the particular app, version, settings and task.

Messengers and social networks

A messenger may process conversations and media, contacts and groups, message delivery details and service information. The operator’s access to content depends on the encryption scheme. For example, Signal’s privacy policy says it cannot read end-to-end encrypted messages but stores technical information needed to run the service. This is the developer’s statement, not our audit.

Do not assume that all social networks work like one messenger. Check message content, backups, technical data and the purposes for which each is used separately.

  1. Check which encryption mode applies to the conversation and backup you need. Set up device locking and available account checks using the service’s documentation.
  2. Save what you need before clearing history. Check which copies are deleted: deletion on your device is not proof of deletion by recipients or the operator.
  3. Review access to contacts and location. Keep permissions needed for the features you use.
Connection testHow fast is your internet right now?
Test your internet?Press to start — takes 30 seconds and up to 50 MB of data
—mspingmax—avg—min——Mbpsdownloadmax—avg—min——Mbpsuploadmax—avg—min—
Jitter—Ping under load—
Last speed test
MetricResultmaxavgmin
Ping—ms———
Download—Mbps———
Upload—Mbps———

Maps and transport services may process route requests, addresses and trips. Not every app automatically creates a continuous location history. For example, Google’s Timeline documentation says the feature is off by default and must be enabled by the user. Location permission, history storage and its backup are separate settings.

  1. Check whether constant location access is needed. Choose access while using the app if that is sufficient; account for features that need background access.
  2. Save what you need before deleting routes and addresses. Check the service’s rules for deleting history and backups.
  3. Separate personal and work trips if it helps organize your data. Different profiles do not by themselves prove that trips cannot be linked.

Email clients and “all-in-one” ecosystems

Email may contain messages, attachments, contacts and notifications from other services. Who can access them depends on the client, email provider, encryption and permissions granted. Signing in with an account does not itself authorize reading the mailbox.

  1. Use a unique password and available two-factor authentication.
  2. Separate addresses for documents and mailing lists if useful. This does not guarantee that the accounts are independent.
  3. Review linked apps to see what is allowed: sign-in, account linking or data access. Google, for example, distinguishes these options. Revoking access may disable features and does not automatically delete data the app has already saved.

Fitness trackers and health apps

Depending on its features, an app may process heart rate, sleep, activity, weight, health records and workout routes. Disclosure can reveal sensitive circumstances of someone’s life.

In February 2023, the FTC brought a case against GoodRx over sharing personal health information with advertising companies contrary to its privacy promises. This concerns unauthorized disclosure alleged by the FTC, not a hack of every fitness app. No comparative ranking of the consequences of leaks is established here.

  1. Check which records the feature needs. Do not enter extra sensitive information just because a field is available.
  2. Read who receives the data, why it is shared, how long it is kept and how to request deletion.
  3. When comparing options, check what stays on the device and what goes to the cloud. The word “local” does not replace checking backups, analytics and data policies.

“Free” utilities and entertainment apps

A utility may need data access for its function: a scanner needs the camera, an editor needs selected images. Check whether the developer explains each requested permission and data sharing with partners. Advertising or an SDK does not by itself prove malicious behavior.

  1. Review installed utilities. Save needed files before removing an unwanted app, and check its subscription and account separately.
  2. Match the permission to the action. If the purpose is unclear, do not grant access until you understand it.
  3. Assess a web version separately if it meets your needs. Before uploading a document or photo, check where the file goes; a browser does not guarantee an absence of data collection.

Why Official App Stores Do Not Guarantee Safety

Official stores apply rules and checks, but their presence does not prove that every app version is safe. Google Play Protect checks apps and warns about potentially harmful behavior. Apple sets requirements for privacy and third-party code.

Before installing, check the developer, the app’s purpose and its data practices. A promise in the listing does not replace checking whether the terms suit you. Do not disable Play Protect to install an app.

How Dangerous Apps Enter Play Market and App Store

After an update or developer change, review changes to features, requested access and data policies. Server settings can also change how a network feature behaves. An update, remote configuration, new owner or delayed execution does not by itself prove malicious behavior.

Review new requests and messages from security tools. Do not disable security updates or device protection because of this article.

On May 23, 2023, ESET described iRecorder — Screen Recorder. The app initially lacked the malicious functionality found later. Code for recording audio and stealing files was subsequently added, most likely in version 1.3.8 released in August 2022. ESET reported that Google Play removed the app after its notification.

This research concerns specific versions. It does not test current apps with similar names or prove that all automatic updates are dangerous.

Permissions: What an App Can Access

Review camera, microphone, contact and location access, as well as SMS and call logs where the system provides it. Available permissions depend on the OS and version. On Android, permission to send notifications does not mean permission to read other apps’ notifications: a notification listener uses a separate access mechanism.

A permission shows possible access, not proof of actual data transfer. Some data is also collected in ways these permissions do not control. Compare permissions, the store listing and the data policy.

SDKs and Third-Party Libraries: Their Role in an App

An SDK is a software development kit. An app may include third-party libraries for advertising, analytics, notifications and installation measurement. Not every SDK sends data to its own server or tracks users. Check the embedded code’s purpose and the developer’s description of its data handling. Apple separately describes disclosures for third-party code.

Which SDK Risks to Check

Check whether data collection, sharing with partners and linking activity across apps are described. Remotely configuring a feature does not itself prove misuse; what it does and whether it follows the stated terms matter. The presence of an SDK does not show how many companies received data. A permissions screen alone cannot verify all embedded code.

What Store Rules Require of Developers

Apple holds developers responsible for third-party code and requires disclosure of data practices. Google Play describes Data safety as information declared by the developer. These requirements help assess an app, but a listing alone does not prove compliance with every rule. Check the particular app’s policy for collection purposes, recipients, retention and deletion.

Fake and Clone Apps in Official Stores

A similar name or icon does not establish a link to the service you need. Apple’s rules explicitly prohibit misleading and copycat names and information.

Before installing, follow the store link from the service’s official website. Compare the developer’s name and the app’s purpose. If they do not match or the origin is unclear, do not enter passwords or payment details until you resolve it.

Why “free” apps are not free

A free download does not explain how an app is funded. Options include advertising, paid features, subscriptions, affiliate commissions, voluntary support or an author working without commercial revenue. Check the particular project. Being free or paid does not by itself prove safety or the sale of data.

Model 1: advertising and behavioral profiling

Advertising may consider a page’s content or information about its audience. These are different ways to select ads. Behavioral personalization may use interests and identifiers; check measurement and retargeting separately under the advertising system’s terms.

Advertising alone does not prove tracking across apps. Review the explanation of personalization and available settings. Further reading: why ads may know you better than friends.

SDKs in Monetization Models

An SDK is a technical way to integrate advertising, analytics and crash reporting, install attribution, fraud prevention and notifications. These functions can support different funding models. A library’s presence alone does not determine whether the author earns revenue or which information is sent. Check the specific purposes and data recipients. More on these checks: dangerous apps and embedded SDKs.

Model 3: data brokerage (direct or indirect)

Check whether an app shares events, identifiers and location with partners, and whether they can link that information to other sources. A dataset without names is not proof that records cannot be linked to a person.

In January 2024, the FTC brought a case against InMarket over location collection and its use for advertising profiles without adequate notice and consent. The FTC described both the company’s own apps and SDKs in third-party apps. This case does not mean that all aggregate statistics can identify someone. Related reading: the risks of centralized data storage.

Model 4: Free Installation and Subscriptions

Installation may be free while some features require a subscription. A trial, an offer to upgrade and automatic renewal do not by themselves prove deception. Before confirming, check available features, the price, trial length, next billing date and cancellation method. If terms are hidden or the interface pressures you, stop and clarify them. More on interface tactics: how data is used to manipulate users.

Model 5: affiliate and commerce monetization

An app may receive a commission for a purchase, another app’s installation, partner registration or the choice of a payment or delivery service. Check whether the commercial link is explained and what data is used to attribute the action. Do not treat a paid placement as an independent recommendation; a commission alone does not prove hidden tracking or malicious behavior.

Why this becomes a security problem, not just a privacy issue

Assess free and paid apps using the same questions:

  • Do permissions match the features you use?
  • Which third-party components and partners receive data?
  • Is the app maintained, and are developer and policy changes clear?
  • What data is retained, and how are access and retention limited?

Data collection affects privacy; credential theft and malicious code execution are separate security threats. One does not automatically prove the other. No comparative study of free and paid apps was conducted here. Related reading: where scammers get your data.

Do not treat uninstalling an app as confirmation that its account and all server copies are deleted. Check data deletion requests and linked app access separately. Google warns that removing a sign-in link does not delete data in the other app; uninstalling an app with a Google Play subscription does not cancel it. Manage subscriptions billed through Apple separately in subscription settings. More: what happens to data after deleting an app.

How to reduce overall app risk

  1. Review installed apps. Choose those you do not need; save needed files and check the account and subscription before removing them.
  2. Review permissions. On Android: Settings → Apps → the app → Permissions. On iPhone: Settings → Privacy & Security → data category → app. Menus depend on the OS version and model; check your device’s documentation for Android or iPhone. Restrict unnecessary access and test the feature you need after changing it.
  3. Separate personal and work tasks where required. Change managed profiles and devices with your administrator; a separate profile does not guarantee that data cannot be linked.
  4. When an incident is reported, open the service’s official incident report and follow instructions for the affected data. Change a compromised password; do not follow a random link in a notification.
  5. Compare the developer and update source. Do not disable protection or security updates for this check.
  6. Before installing, read the store’s data information and the policy. After installation, compare them with permission requests. Stop and clarify the purpose of access you do not understand.

Keep apps whose functions you need and whose data practices suit your task. If an access purpose or data recipient is unclear, consult the documentation first. Reviewing permissions helps restrict access but does not establish the safety of all the code.

Author: — Cybersecurity specialist, VPNON