Headline
New Study Warns Several Free iOS and Android VPN Apps Leak Data
A Zimperium zLabs analysis of 800 free Android and iOS VPN apps exposes critical security flaws, including the Heartbleed bug, excessive system permissions, and non-transparent data practices. Learn how these ‘privacy’ tools are actually major security risks, especially for BYOD environments.
Millions who rely on free mobile Virtual Private Network (VPN) apps for online privacy may actually be putting their data at greater risk, according to new research by Zimperium zLabs. In a study of nearly 800 free VPN apps for Android and iOS, researchers found many not only fail to protect users but also expose them to serious security and privacy threats.
****Critical Flaws Discovered:****
The zLabs team discovered that a substantial portion of these apps exhibit dangerous behaviours. Some leak personal data, while many others offer “no real privacy at all.” Researchers noted a major concern is the developers’ use of incredibly old and vulnerable software.
For example, the analysis found three VPN apps still use an outdated part of the OpenSSL library, leaving them open to the infamous Heartbleed bug (CVE-2014-0160). This flaw, revealed in 2014, could allow a remote attacker to read sensitive information like secret keys, usernames, and passwords.
About 1% of the apps were vulnerable to Man-in-the-Middle (MitM) attacks, giving attackers the ability to intercept and read all user traffic. Releasing an app with a decade-old flaw that has a known fix highlights a serious lack of security diligence.
****Excessive Permissions and Surveillance:****
Further probing revealed that many apps are also requesting powerful, unnecessary access, a practice known as Permission Abuse. For instance, an iOS VPN app asking for “always-on” location access (LOCATION_ALWAYS) makes no sense, since a VPN’s main job is to secure traffic, not track your physical location 24/7.
Source: Zimperium
Similarly, some Android apps requested the ability to read all system logs (READ_LOGS), which could allow them to build a full profile of a user’s behaviour, thereby operating as a “sophisticated keylogger.”
Some apps asked for permissions like access to microphones, system logs, or performed UI screen capture, giving the app provider a surveillance vector well beyond its stated function.
****Non-Transparent Privacy Practices:****
According to Zimperium zLabs’ blog post, researchers found a prevalent lack of transparency among their inspected apps, hindering users’ ability to give informed consent about the data being collected. Even on Apple’s App Store, a massive 25% of iOS VPN apps lacked a valid privacy manifest, a core requirement meant to inform users how their data will be handled.
Source: Zimperium
Additionally, over 6% of these iOS apps requested private entitlements, which are powerful permissions that could allow deep access to the operating system and should never be available to third-party developers.
For companies that let staff use their personal devices for work (called Bring-Your-Own-Device or BYOD policies), these insecure VPNs can become the weakest link, putting sensitive business data at unnecessary risk. Ultimately, when it comes to free mobile VPNs, what is assumed to be protecting your privacy may actually be the biggest risk to your data.
“Organizations need a multi-layered response. Endpoint visibility and management is table stakes. Some organizations will evaluate the risk and tackle this through application allow listing, while others may favor a more permissive approach. However, what is rapidly becoming a requirement is the need for web content-level data security,“ said Brandon Tarbet, Director, IT & Security at Menlo Security.
“This need is underscored by how personal VPN providers position and market the supposed security benefits of their products,“ Tarbet warned. “There is a real need for data protection at the content level, and a market that wants to be able to trust their connection to websites and services. The key is shifting from a perimeter-based security mindset (such as with VPNs) to content-level protection that works even when traditional visibility is compromised,” he urged.
Related news
In Bitcoin Core through 26.0 and Bitcoin Knots before 25.1.knots20231115, datacarrier size limits can be bypassed by obfuscating data as code (e.g., with OP_FALSE OP_IF), as exploited in the wild by Inscriptions in 2022 and 2023.
Red Hat strives to get better at what we do, faster at how we do it, while maintaining high quality results. In modern software development, that means focusing on security as early as possible into our software development process, and continuously driving improvements by listening and acting upon early feedback in the Secure Development Lifecycle (SDL). One important tool toward that goal is the Common Weakness Enumeration (CWE), a community-developed taxonomy of flaws. We use CWE classifications to gather intelligence and data to visualize clustering common weaknesses. We can then better
Organizations should update to the latest encryption (version 3.0.7) as soon as possible, but there's no need for Heartbleed-like panic, security experts say.
The OpenSSL project has rolled out fixes to contain two high-severity flaws in its widely used cryptography library that could result in a denial-of-service (DoS) and remote code execution. The issues, tracked as CVE-2022-3602 and CVE-2022-3786, have been described as buffer overrun vulnerabilities that can be triggered during X.509 certificate verification by supplying a specially-crafted email
Zimbra Collaboration Open Source 8.8.15 does not encrypt the initial-login randomly created password (from the "zmprove ca" command). It is visible in cleartext on port UDP 514 (aka the syslog port).
Under certain circumstances, a vulnerability in Metasys ADS/ADX/OAS 10 versions prior to 10.1.5 and Metasys ADS/ADX/OAS 11 versions prior to 11.0.2 could allow a user to inject malicious code into the MUI Graphics web interface.
Multiple buffer overflows in Active Management Technology (AMT) in Intel Manageability Engine Firmware 8.x/9.x/10.x/11.0/11.5/11.6/11.7/11.10/11.20 allow attacker with local access to the system to execute arbitrary code with AMT execution privilege.
Unspecified vulnerability in Oracle Java SE 6u75, 7u60, and 8u5 allows remote attackers to affect integrity via unknown vectors related to Deployment.
Unspecified vulnerability in the MySQL Server component in Oracle MySQL 5.5.37 and earlier, and 5.6.17 and earlier, allows remote authenticated users to affect integrity and availability via vectors related to SRCHAR.
The (1) TLS and (2) DTLS implementations in OpenSSL 1.0.1 before 1.0.1g do not properly handle Heartbeat Extension packets, which allows remote attackers to obtain sensitive information from process memory via crafted packets that trigger a buffer over-read, as demonstrated by reading private keys, related to d1_both.c and t1_lib.c, aka the Heartbleed bug.