There has been a 30% increase in mobile malware in Fortinet’s laboratories over the last six months, according to the FortiGuard threat landscape research for the first half of 2013.

The team is now seeing more than 1 300 new samples per day, is currently tracking over 300 unique Android malware families and more than 250 000 unique malicious Android samples.

The BYOD phenomenon has many benefits for a business, chief among them are increased employee efficiency and productivity gains. However, the disadvantage of a lenient BYOD policy is the threat of mobile malware infecting the user’s device and, subsequently, the business network.

“Three years ago, mobile malware wasn’t much of a concern for users or businesses. Most malware at the time targeting smartphones and tablets were nothing more than annoyware such as the Cabir virus or scam software used to commit SMS fraud or replace icons,” says Axelle Apvrille, senior mobile anti-virus researcher for Fortinet’s FortiGuard Labs.

“However, as devices have proliferated, so, too, have cybercriminals eager to capitalize on the growing user base, and our research shows the proliferation of mobile malware will not abate anytime soon.”

In 2009, the majority of mobile malware in existence targeted the Symbian OS, as iOS and Android were still relatively new in the marketplace. In addition, a large number of the malware was coded by programmers in Eastern Europe and China, places where Symbian commanded a large share of the user base.

In 2013, the mobile threat landscape changed dramatically. Wide scale manufacturer adoption of Google’s Android OS globally has led to an explosion of smartphones in the marketplace.

Android devices are available in every market, at price levels from the incredibly inexpensive to feature-rich, cutting edge computing monsters. Coupled with the explosion of available applications to extend device functionality, cybercriminals and other nefarious types have used this platform as a new business opportunity.

In 2012, FortiGuard predicted that the financially lucrative ransomware would make its way on to mobile phones.

“Ransomware has been incredibly successful financially for cybercriminals, it’s no surprise they’ve turned their attention to mobile devices,” says Richard Henderson, security strategist for Fortinet’s FortiGuard Labs. “The Fake Defender malware for Android follows the same MO as PC fake antivirus software – it pretends to be altruistic, but in reality, it lies in wait to launch its true form.

“This malware then locks the victim’s phone and demands payment before unlocking the device. Once the phone is locked, the victim can either pay the ransom or completely erase their device, losing all their photos and data unless they have a full backup elsewhere.”

Even though we’ve seen recent patches for Ruby on Rails, Java, Adobe Acrobat and Apache, FortiGuard Labs is finding attackers are still exploiting those old vulnerabilities.

In January, it was announced that a critical vulnerability in the Ruby on Rails Framework could allow a remote attacker to execute code on the underlying Web server.

Ruby on Rails (RoR) is a Web application framework for the Ruby programming language. Put simply, it allows for rapid, easy and elegant deployment of “Web 2.0” Websites. RoR is a popular framework: hundreds of thousands of Websites online use RoR in some fashion.

Further adding to the problem, a Metasploit module was made available to scan for the vulnerability, making the ability to find a Web server to exploit a trivial matter.

“The exploit involved a flaw in the XML processor deserialisation routine, which is used to create Ruby objects on the fly,” says Henderson. “RoR was patched to correct the flaw, but four months later it was discovered that an attacker or attackers was searching for and exploiting, unpatched Web servers in order to infect them with software.”

In January, a zero-day exploit that was able to bypass Java’s sandbox and run arbitrary Java code was discovered.

Java is a ubiquitous technology online – most computers have some form of Java installed and enabled. The vulnerability allowed a malicious applet to run any Java program, bypassing Java’s sandbox and granting full access to the vulnerable computer.

Attacks were discovered in the wild and the exploit was quickly integrated into many popular crimeware attack kits, such as BlackHole, Redkit and Nuclear Pack, giving purchasers of these kits the ability to take advantage of the exploit and install malware on computers. A Metasploit module was also created for the vulnerability, making the ability to find victims a simple point and click affair.

“The exploit involved a flaw in a JMX (Java Management Extensions) component that allowed the malicious applet to elevate its privileges and run any Java code it wished,” Henderson says.

Oracle was quick to release a patch for the flaw – but similar to other exploits integrated into crimeware kits, many new victims were found – and continue to be found – running unpatched versions of Java, allowing malware to be installed.

In February, a PDF pretending to be a travel visa form from Turkey was detected circulating in the wild and took advantage of a previously unseen vulnerability in Adobe’s Reader software. The exploit worked with all recent versions of Adobe Reader (9.5.X, 10.1.X, and 11.0.X), and on most versions of Microsoft Windows, including 64-bit Windows 7 and most Mac OS X systems.

The exploit PDF was used by cybercriminals in order to install malware on their target’s computers.

Adobe released a patch for the flaws in Reader on February 20th, but cybercriminals continue to use repackaged versions of the exploits online in spear-phishing attacks. These exploits in Adobe’s Reader software continue to be actively exploited by cybercriminals as a malware delivery method, taking advantage of users who don’t expeditiously patch their computers.

In late April, a new attack on the popular Apache Web server was discovered. Dubbed CDorked, the malware was able to compromise the Web server and redirect visitors of the compromised Web server to other servers that deliver malware using the BlackHole exploit kit. The attack may also have targeted the Lighttpd and Nginx Web server platforms.

CDorked shows many similarities to 2012’s DarkLeech attack on Apache servers, but is significantly stealthier and smarter than DarkLeech was: unlike DarkLeech, CDorked didn’t load additional malicious modules on the infected server; instead it maliciously modified the existing httpd binary.

CDorked was interesting in that it did not write any information to the Web server’s hard drive: everything was kept in memory and was accessed via obfuscated GET requests sent by the attackers to the compromised server. None of those GET requests were logged.
CDorked showed some intelligence in how it operated.
“It had a quota system built in,” Henderson says. “In other words, CDorked did not attempt to redirect each and every visitor to a BlackHole site. It also hid from users attempting to access administrative pages on the compromised Web server in an attempt to keep users who may have been more likely to notice a redirect to a crimeware delivery site from discovering the compromise.
“CDorked isn’t alone though: other malicious malware have intelligence built-in to watch for malware analysts and other White Hat hackers.”