Security researchers are watching a new botnet, GoldBrute, which is currently brute-forcing a list of roughly 1.5 million remote desktop protocol (RDP) servers exposed online. The ongoing campaign is one of many scanning for vulnerable servers and using weak or reused passwords to access them.
RDP has been making headlines since Microsoft disclosed "BlueKeep," a remote code execution vulnerability that includes RDP in its attack chain. But botnets have been hunting vulnerable RDP servers for years, explains Renato Marinho, chief research officer at Morphus Labs, in a blog post. GoldBrute uses its own list, which it continues to build as it scans for credentials.
The botnet initially scans for random IP addresses to find hosts with exposed RDP servers; these addresses are sent back to the command-and-control (C2) server. Once the bot reports 80 victims, the C2 assigns it a set of targets to brute-force. Each bot only tries one specific username and password, which researchers expect is a strategy to bypass security tools.
After the RDP target is successfully brute-forced, it downloads a large zip file with the GoldBrute Java code and Java runtime then runs a jar file ("bitcoin.dll"). The new bot then starts to scan for open RDP servers they call "brutable," and these are sent to the C2 via WebSocket connection. When the bot reaches 80 brutable servers, it begins the brute-force phase: This means the bot continuously receives and brute-forces "host + username + password" combos," which are passed along to the attackers to continue building the list of valid combinations.
Read more details here.