A study led by researchers from Purdue University has found that a notable number of Android applications marketed toward U.S. military personnel include software components developed by companies based in China, Russia, and other countries identified by the U.S. Department of Defense as adversarial nations.
The research, conducted in collaboration with the U.S. Military Academy at West Point and Florida International University, examined more than 220 Android applications obtained from Google Play and online military communities. According to the researchers, over one in every eight apps analyzed contained code associated with organizations headquartered in countries regarded as strategic competitors of the United States. The Pentagon declined to comment on the study's findings.
Rather than identifying malicious applications outright, the study highlights a growing software supply chain challenge created by third-party Software Development Kits (SDKs), which developers routinely integrate into applications to support features such as advertising, analytics, authentication, and push notifications. While these components simplify development, they can also introduce external code that developers may not fully inspect or even realize has been included.
Researchers found SDKs in approximately 64% of the analyzed applications. Among them, twelve apps contained Huawei's HMS Core framework, including applications developed for state National Guard organizations. In one instance, Huawei's software was not intentionally added by the application's developer. Instead, it was introduced indirectly through a commercial notification service that bundled Huawei's SDK as a dependency.
The researchers noted that this type of indirect integration presents an important security concern because SDKs can receive remote updates over time. Even if no sensitive information is transmitted today, future updates could potentially alter an application's behavior without users being aware. Although the study did not observe any data being sent to Huawei-controlled servers during testing, the researchers cautioned that the presence of such software should not automatically be considered harmless.
Beyond Huawei, the analysis also identified software associated with Russian technology companies. SDKs linked to Yandex advertising services were found in applications used by military-affiliated users. The study also referenced Pushwoosh, a Russian software company that previously presented itself as a U.S.-based business. Reuters reported in 2022 that Pushwoosh code had been embedded in official mobile applications operated by the U.S. Army and the U.S. Centers for Disease Control and Prevention (CDC). Both organizations subsequently removed the software following public disclosure.
The researchers also discovered discrepancies between application behavior and the privacy information disclosed through Google Play. Approximately 40% of the applications examined either collected or shared more user data than indicated by their app store privacy labels. Overall, around 7% of the analyzed applications contained software linked to companies headquartered in countries designated by the Pentagon as adversarial nations.
Lead author Joshua Shinkle of Purdue University said the team hopes the findings encourage stronger awareness among military personnel, application developers, platform providers, and policymakers. According to Shinkle, the research is intended to support more informed privacy decisions while encouraging discussions about improving transparency and addressing existing security gaps.
The study argues that the implications extend beyond software development practices. Researchers pointed to the commercial mobile advertising ecosystem, where applications routinely collect location and device information that can later be shared through data brokers. Such information has the potential to expose sensitive operational patterns, including troop movements, deployment routines, and activity around military installations.
The report references an April letter in which U.S. Central Command (CENTCOM) informed Senator Ron Wyden that it had received multiple threat reports indicating that adversaries were exploiting commercially available location data to monitor U.S. military personnel operating near Iran and the Strait of Hormuz. Lawmakers described the disclosure as the first official acknowledgement that commercially traded mobile data had been used to track troops deployed in an active conflict zone.
Researchers also surveyed 103 military-affiliated Americans to better understand attitudes toward mobile application privacy. More than 83% of respondents reported using at least one application whose data collection practices made them uncomfortable. Between 76% and 83% indicated they would be extremely uncomfortable using applications containing software developed by companies from adversarial nations.
Despite these concerns, participants expressed greater trust in applications carrying military branding, suggesting that official appearance can influence perceptions of security even when underlying software components remain largely invisible to users. Nearly two-thirds of respondents also reported receiving little or no institutional guidance regarding the security risks associated with personal mobile applications.
When asked about potential solutions, respondents strongly supported greater transparency regarding third-party software embedded within applications. The most widely supported recommendation involved providing users with in-device notifications identifying foreign-developed SDKs before installation or use.
Participants also backed stronger federal restrictions on the commercial trading of military-affiliated location data, independent security audits of applications, and tighter controls on the inclusion of foreign-developed SDKs in apps marketed toward service members.
While the U.S. Marine Corps already prohibits several categories of applications, including gambling, dating, and cryptocurrency apps, from government-issued devices and has warned personnel against using platforms such as TikTok and WeChat, the researchers argue that personal smartphones remain a largely unaddressed area of risk because they frequently fall outside existing policy controls.
According to the researchers, improving software transparency will require greater visibility into third-party components that operate behind the scenes. They recommend clearer country-of-origin labeling for embedded SDKs within application marketplaces and encourage developers to regularly audit their software dependency chains to better understand which external components are included in their applications.
The study concludes that branding alone should not be viewed as an indicator of application security. As modern mobile apps increasingly rely on extensive networks of third-party software, researchers argue that stronger transparency, routine dependency auditing, and more robust privacy safeguards will be necessary to reduce hidden supply chain risks facing military personnel and other users handling sensitive information.
Google has formally begun rolling out a comprehensive verification framework for Android developers, a move aimed at tackling the persistent problem of malicious applications being distributed by actors who operate without revealing their identity. The company’s decision reflects growing concerns within the mobile ecosystem, where anonymity has often enabled bad actors to bypass accountability and circulate harmful software at scale.
This rollout comes in advance of a stricter compliance requirement that will first take effect in September across key markets including Brazil, Indonesia, Singapore, and Thailand. These regions are being used as initial enforcement zones before the policy is gradually expanded worldwide next year, signaling Google’s intent to standardize developer accountability across its global Android ecosystem.
Under the new system, developers who distribute Android applications outside of the official Google Play marketplace will now be required to register through the Android Developer Console and verify their identity credentials. This requirement is particularly substantial for developers who rely on alternative distribution methods such as direct APK sharing, enterprise deployment, or third-party app stores, as it introduces a layer of traceability that previously did not exist.
At the same time, Google clarified that developers already publishing applications through Google Play and who have completed existing identity verification processes may not need to take further action. In such cases, their applications are likely to already comply with the updated requirements, reducing friction for those operating within the official ecosystem.
Explaining how this change will affect end users, Matthew Forsythe, Director of Product Management for Android App Safety, emphasized that the vast majority of users will not notice any difference in their day-to-day app installation experience. Standard app downloads from trusted sources will continue to function as usual, ensuring that usability is not compromised for the general public.
However, the experience changes when a user attempts to install an application that has not been registered under the new verification system. In such cases, users will be required to proceed through more advanced installation pathways, such as Android Debug Bridge or similar technical workflows. These methods are typically used by developers and experienced users, which effectively limits exposure for less technical individuals.
This design introduces a deliberate separation between general users and advanced users. While everyday users are shielded from potentially unsafe applications, power users retain the flexibility to install software manually, albeit with additional steps that reinforce intentional decision-making.
To further support developers, Google is integrating visibility into its core development tools. Within the next two months, developers using Android Studio will be able to directly view whether their applications are registered under the new system at the time of generating signed App Bundles or APK files. This integration ensures that compliance status becomes part of the development workflow rather than a separate administrative task.
For developers who have already completed identity verification through the Play Console, Google will automatically register eligible applications under the new framework. This automation reduces operational overhead and ensures a smoother transition. However, in cases where applications cannot be automatically registered, developers will be required to complete a manual claim process to verify ownership and bring those apps into compliance.
In earlier guidance, Google also outlined how sideloading, the practice of installing apps from outside official stores, will function under this system. Advanced users will still be able to install unregistered APK files, but only after completing a multi-step verification process designed to confirm their intent.
This process includes an authentication step to verify the user’s decision, followed by a one-time waiting period of up to 24 hours. The delay is not arbitrary. It is specifically designed to disrupt scam scenarios in which attackers pressure users into quickly installing malicious applications before they have time to reconsider.
Forsythe explained that although this process is required only once for experienced users, it has been carefully structured to counter high-pressure social engineering tactics. By introducing friction into the installation process, the system aims to reduce the success rate of scams that rely on urgency and manipulation.
This development is part of a wider industry tendency toward tightening control over app ecosystems and improving user data protection. In a parallel move, Apple has recently updated its Developer Program License Agreement to impose stricter rules on how third-party wearable applications handle sensitive data such as live activity updates and notifications.
Under Apple’s revised policies, developers are explicitly prohibited from using forwarded data for purposes such as advertising, user profiling, training machine learning models, or tracking user location. These restrictions are intended to prevent misuse of real-time user data beyond its original functional purpose.
Additionally, developers are not allowed to share this forwarded information with other applications or devices, except for authorized accessories that are explicitly approved within Apple’s ecosystem. This ensures tighter control over how data flows between devices.
The updated agreement also introduces further limitations. Developers are barred from storing this data on external cloud servers, altering its meaning in ways that change the original content, or decrypting the information anywhere other than on the designated accessory device. These measures collectively aim to preserve data integrity and minimize the risk of misuse.
Taken together, this charts a new course across the technology industry toward stronger governance of developer behavior, application distribution, and data handling practices. As threats such as malware distribution, financial fraud, and data exploitation continue to evolve, platform providers are increasingly prioritizing transparency, accountability, and user protection in their security strategies.
Various mental health mobile applications with over millions of downloads on Google Play have security flaws that could leak users’ personal medical data.
Researchers found over 85 medium and high-severity vulnerabilities in one of the apps that can be abused to hack users' therapy data and privacy.
Few products are AI companions built to help people having anxiety, clinical depression, bipolar disorder and stress.
Six of the ten studied applications said that user chats are private and encoded safely on the vendor's servers.
Oversecured CEO Sergey Toshin said that “Mental health data carries unique risks. On the dark web, therapy records sell for $1,000 or more per record, far more than credit card numbers.”
Experts scanned ten mobile applications promoted as tools that help with mental health issues, and found 1,575 security flaws: 938 low-severity, 538 medium-severity, and 54 rated high-severity.
No critical issues were found, a few can be leveraged to hack login credentials, HTML injection, locate the user, or spoof notifications.
Experts used the Oversecured scanner to analyse the APK files of the mental health apps for known flaw patterns in different categories.
Using Intent.parseUri() on an externally controlled string, one treatment app with over a million downloads launches the generated messaging object (intent) without verifying the target component.
This makes it possible for an attacker to compel the application to launch any internal activity, even if it isn't meant for external access.
Oversecured said, “Since these internal activities often handle authentication tokens and session data, exploitation could give an attacker access to a user’s therapy records.”
Another problem is storing data locally that gives read access to all apps on the device. This can expose therapy details, depending on the saved data. Therapy details such as Cognitive Behavioural Therapy (CBT), session notes, therapy entries. Experts found plaintext configuration data and backend API endpoints inside the APK resources.
“These apps collect and store some of the most sensitive personal data in mobile: therapy session transcripts, mood logs, medication schedules, self-harm indicators, and in some cases, information protected under HIPAA,” Oversecured said.
In recent developments, threat actors are using a technique known as "versioning" to evade Google Play Store's malware detection mechanisms, posing a significant risk to Android users. This method allows them to specifically target users and compromise their sensitive information, including credentials, data, and finances. Despite being a known tactic, versioning remains challenging to detect, making it a preferred choice for malicious developers.
In a blog post published on April 10, Kaspersky researchers reported their findings of a thorough analysis of nine of the most well-known Dark Web forums. They discovered a booming market of buyers and sellers exchanging access to botnets, malicious Android applications, and app developer accounts for hundreds of dollars at a time by monitoring activities between 2019 and 2023.
Some highly valuable products, such as source code that can let a threat actor hack into an existing cryptocurrency or a dating app on Google Play can cost several thousand dollars.
"It's an infinite cat and mouse game[…]The attackers find a way to bypass security scanners. Then the people developing the security scanners deploy patches to ensure that doesn't happen again. Then the attackers find new flaws. And it goes on and on," says Georgy Kucherin, Kaspersky research with regards to Google’s app security.
Any program that is posted to the Apple or Google app stores undergoes a rigorous inspection. However, according to the Kaspersky researchers “just like any security solution that exists in the world, it's not 100% effective[…]Every scanner contains flaws that threat actors exploit to upload malware to Google Play."
Commonly, there are two methods by with a hacker attempts to sneak malware onto an app store:
Moreover, depending on the developer, access to a Google Play account may only cost as little as $60, depending on the developer. However, other, more beneficial accounts, resources, and services have significantly greater costs.
For example, considering the power they hold, loaders — the software necessary to deploy malicious code into an Android app — can cost big bucks on the darknet markets, ranging up to a whopping $5,000 each for an instance.
A well-resourced criminal could well go with a premium package, like the source code for a loader.
"You can do whatever you want with that — deploy it to as many apps as you want[…]You can modify the code as much as you want, adapting it to your needs. And the original developer of the code may even provide support, like updates for the code, and maybe new ways to bypass security measures," Kucherin explains.
The threats posed by Google Play are a cause of great concern to organizations, especially the ones with feeble enterprise security. Kucherin notes that many businesses still have lax bring-your-own-device arrangements in place, which extend the security perimeter outside of corporate networks and right into the hands of its employees.
"Say an employee installs a malicious app on the phone[…]If this app turns out to be a stealer, cybercriminals can get access to, for example, corporate emails or sensitive corporate data, then they can upload it to their servers and sell it on the Dark Web. Or even worse: An employee might keep their passwords in, for example, their phone's notes app. Then hackers can steal those notes and get access to corporate infrastructure," he explains.
In order to prevent such severe outcomes, Kucherin suggests two simple precautionary measures:
One, you can teach the employees cyber-hygiene principles, like not downloading apps that are not trusted. However, this might not suffice, so "another thing you can do — though it's more expensive — is give your employees a separate phone, which they will use only for purposes of work. Those devices will contain a limited number of apps — just the essentials like email, phone, no other apps allowed,” he adds.
Just as it is for the cybercriminals, you have to pay more to get more, he notes: "Using dedicated work devices is more effective, but more expensive."
The study was conducted between the 40 most downloaded Android apps, out of which 20 were free apps and 20 were paid, on Google Play and found that nearly 80% of these apps disclose misleading or false information.
The following findings were made by the Mozilla researchers:
Google apparently launched its data privacy section for the Play Store last year. This section was introduced in an attempt to provide a “complete and accurate declaration” for information gathered by their apps by filling out the Google Data Safety Form.
Due to certain vulnerabilities in the safety form's honor-based system, such as ambiguous definitions for "collection" and "sharing," and the failure to require apps to report data shared with "service providers," Mozilla claims that these self-reported privacy labels may not accurately reflect what user data is actually being collected.
In regards to Google’s Data Safety labels, Jen Caltrider, project lead at Mozilla says “Consumers care about privacy and want to make smart decisions when they download apps. Google’s Data Safety labels are supposed to help them do that[…]Unfortunately, they don’t. Instead, I’m worried they do more harm than good.”
In one instance in the report, Mozilla notes that TikTok and Twitter both confirm that they do not share any user data with the third parties in their Data Safety Forms, despite stating that the data is shared with the third parties in their respective privacy policies. “When I see Data Safety labels stating that apps like Twitter or TikTok don’t share data with third parties it makes me angry because it is completely untrue. Of course, Twitter and TikTok share data with third parties[…]Consumers deserve better. Google must do better,” says Caltrider.
In response to the claim, Google has been dismissing Mozilla’s study by deeming its grading system inefficient. “This report conflates company-wide privacy policies that are meant to cover a variety of products and services with individual Data safety labels, which inform users about the data that a specific app collects[…]The arbitrary grades Mozilla Foundation assigned to apps are not a helpful measure of the safety or accuracy of labels given the flawed methodology and lack of substantiating information,” says a Google spokesperson.
Apple, on the other hand, has also been criticized for its developer-submitted privacy labels. The 2021 report from The Washington Post indicates that several iOS apps similarly disclose misleading information, along with several other apps falsely claiming that they did not collect, share, or track user data.
To address these issues, Mozilla suggests that both Apple and Google adopt an overall, standardized data privacy system across all of their platforms. Mozilla also urges that major tech firms shoulder more responsibility and take enforcement action against apps that fail to give accurate information about data sharing. “Google Play Store’s misleading Data Safety labels give users a false sense of security[…]It’s time we have honest data safety labels to help us better protect our privacy,” says Caltrider.
![]() |
Users complain that the app doesn't work and just deletes itself |