Cybercriminals have developed a Web-based attack tool to hijack routers on a large scale when users visit compromised websites or view malicious advertisements in their browsers.
The goal of these attacks is to replace the DNS (Domain Name System) servers configured on routers with rogue ones controlled by attackers. This allows hackers to intercept traffic, spoof websites, hijack search queries, inject rogue ads on Web pages and more.
The DNS is like the Internet's phonebook and plays a critical role. It translates domain names, which are easy for people to remember, into numerical IP (Internet Protocol) addresses that computers need to know to communicate with each other.
The DNS works in a hierarchical manner. When a user types a website's name in a browser, the browser asks the operating system for that website's IP address. The OS then asks the local router, which then queries the DNS servers configured on it — typically servers run by the ISP. The chain continues until the request reaches the authoritative server for the domain name in question or until a server provides that information from its cache.
If attackers insert themselves in this process at any point, they can respond with a rogue IP address. This will trick the browser to look for the website on a different server; one that could, for example, host a fake version designed to steal the user's credentials.
An independent security researcher known online as Kafeine recently observed drive-by attacks launched from compromised websites that redirected users to an unusual Web-based exploit kit that was specifically designed to compromise routers.
The vast majority of exploit kits sold on underground markets and used by cybercriminals target vulnerabilities in outdated browser plug-ins like Flash Player, Java, Adobe Reader or Silverlight. Their goal is to install malware on computers that don't have the latest patches for popular software.
The attacks typically work like this: Malicious code injected into compromised websites or included in rogue ads automatically redirect users' browsers to an attack server that determines their OS, IP address, geographical location, browser type, installed plug-ins and other technical details. Based on those attributes the server then selects and launches the exploits from its arsenal that are most likely to succeed.
The attacks observed by Kafeine were different. Google Chrome users were redirected to a malicious server that loaded code designed to determine the router models used by those users and to replace the DNS servers configured on the devices.
Many users assume that if their routers are not set up for remote management, hackers can't exploit vulnerabilities in their Web-based administration interfaces from the Internet, because such interfaces are only accessible from inside the local area networks.
Sign up for Computerworld eNewsletters.