DoubleDoor, a new IoT Botnet bypasses firewall using two backdoor exploits
14.2.2018 securityaffairs Exploit  IoT

Security researchers spotted a new IoT botnet dubbed DoubleDoor that is able to bypass firewall as well as modem security using two backdoor exploits.
IoT devices continue to be a privileged target of cyber criminals, cyber attackers against so-called smart objects has seen a rapid evolution. Security researchers at NewSky Security (NewSky Security) have detected a new IoT botnet dubbed DoubleDoor that is able to bypass firewall as well as modem security using two backdoor exploits.

The analysis of the honeypot logs allowed the researchers to detect the new threat, it leverages two known backdoor exploits to manage two levels of authentications.

The first malicious code is the Juniper Networks SmartScreen OS exploit, it triggers the flaw CVE-2015–7755 to bypass the firewall authentication.

CVE-2015–7755 hardcoded backdoor affects the Juniper Networks’ ScreenOS software that powers their Netscreen firewalls.

“Essentially the telnet and SSH daemons of Netscreen firewalls can be accessed by using the hardcoded password <<< %s(un=’%s’) = %u with any username, regardless of it being valid or not.We saw its implementation in the initial attack cycle of DoubleDoor as it attacked our honeypots with username “netscreen” and the backdoor password.” wrote Ankit Anubhav, Principal Researcher, NewSky Security.

Once succeeded, the malicious code uses the CVE-2016–10401 Zyxel modem backdoor exploit to take full control over the IoT device.

The code is a privilege escalation exploit, “which is why the DoubleDoor attackers also performed a password based attack to get a basic privilege account like admin:CenturyL1nk before going for the superuser.”

“This time it was CVE-2016–10401 , a backdoor for ZyXEL PK5001Z devices. This backdoor is straight forward too, with a hardcoded su password as zyad5001.” continues the expert.

The experts highlighted that differently from other IoT botnets like Satori or Masuta, the DoubleDoor botnet doesn’t use a unique string in the reconnaissance phase.

“after the threat actors have performed the attack, they want a confirmation whether they were successful of getting control of the IoT device. For this, they try to invoke the shell with invalid commands. If the attacker has succeeded, it will show “{string}: applet not found” where {string} is the invalid command.” observed the research.

“DoubleDoor botnet takes care of this, by using a randomized string in every attack”

The DoubleDoor botnet seems to be in an early stage, most of the attacks are originated from South Korean IPs.

The botnet includes the code to target a limited number of devices, it will succeed only if the victim has a specific unpatched version of Juniper ScreenOS firewall which protects unpatched Zyxel modems.

“Double layer of IoT protection is more common in corporate environments, which don’t rely on built-in IoT authentication and like to protect it with another layer of firewall. Although such corporate devices can be lesser in number, getting control of corporate environment routers can be more valuable for an attacker as it can lead to targeted IoT attacks.” concluded the experts.