There has been increased awareness regarding the attack surface inherent in remote desktop protocol (RDP) since both Microsoft and the National Security Agency have recently come out in the strongest possible terms advising Windows users to patch against the BlueKeep (CVE-2019-0708) exploit.
Renato Marinho of Morphus Labs took a look at what was currently transpiring on RDP networks and posted what he found to the SANS ISC InfoSec Forums.
He found that botnets have been scanning for RDP servers and are using weak and reused passwords to gain access to them. This kind of attack is called "brute force," for obvious reasons.
What he calls "GoldBrute" is a botnet which is currently brute-forcing a list of about 1.5 million RDP servers that have been exposed to the Internet. Shodan lists about 2.4 million Internet-exposed servers, in all. But GoldBrute makes its own list of server addresses to attack, which is expanding as the bot continues to scan and grow.
There is one single command and control center (C&C) running GoldBrute. Bot members communicate with it through AES encrypted WebSocket connections on port 8333.
Once infected, the victim system downloads the 80 megabytes of bot code which includes the complete Java Runtime.
When established, the bot will start looking for way to spread itself. It will scan seemingly random IP addresses, while reporting the results of a login attempt back to the C&C machine.
One of the sophistications of GoldBrute is that each bot will only try one particular username and password per target. It does not repeat this action multiple time because that kind of behavior could set off the victim's detection alarms.
The flow of the botnet's handling of data is unique to it. For example, it is only after 80 successful infections have been carried out when the actual "brute force" phase of the attack begins. The bot element will get list of bruting targets from the C&C.
In the brute-force phase, the bot will continually receive and brute-force "host + username + password" combinations.
Marinho did some fiddling and got the botnet to put all the "brutable" IP addresses it looked at onto his own research system. For only one six-hour period, he got 2.1 million IP addresses from the C2 server of which 1,596,571 were unique. This is one serious botnet.
Brute-force attacks are currently the method of choice for cybercriminals trying to leverage RDP. But they can be simply defended against by changing some default setups. It's simple to do on an individual machine -- it only takes one line. An enterprise will find that the one line of change would need to be repeated across the entire network.
— Larry Loeb has written for many of the last century's major "dead tree" computer magazines, having been, among other things, a consulting editor for BYTE magazine and senior editor for the launch of WebWeek.