Damballa, which provides anti-botnet services for enterprises, today revealed the Top 10 botnets it found in its customers' networks in 2009; the so-called ZeusBotnet accounted for nearly 20 percent of all bot infections, while the KoobfaceBotnetB botnet accounted for 15 percent.
Koobface overall had a surprisingly large representation. The worm, typically spread via social networks such as Facebook and MySpace, was the main malware carried by two additional botnets, Koobface.D (5 percent) and Koobface.C (4 percent). The malware was used as a foot in the door to hijack corporate users' accounts and to spread among other systems within the organization, according to Gunter Ollmann, vice president of research for Damballa.
Koobface also was the most common type of malware family used by all botnets to infect enterprises, with Zeus a close second, according to Ollmann.
Meanwhile, a click-fraud botnet known as ClickFraudBotnet was behind 9 percent of enterprise botnet infections, followed by SpamFraudBotnet with 8 percent, both of which staked out the enterprise machines to do their bidding for financial gain, rather than for stealing anything from the victim organizations.
ClickFraudBotnet, for instance, apparently chose its enterprise victims so it could operate under the radar with large amounts of bandwidth and when users weren't on their machines. "They focused on the enterprise for a homogenous infrastructure. We have seen this botnet operation focused on an enterprise for an ease-of-management perspective," he says.
Enterprise bots are especially valuable for click fraud and spamming purposes because they come with IP addresses with good reputations, he says, which means they are most likely to pass anti-spam gateways using IP reputation filters.
ConfickerBotnetA was No. 10 on the list of largest botnet outbreaks, spreading mainly via USB drives, Ollmann says. The Conficker malware family ranked No. 5 in the top 10 botnet malware families discovered in enterprises.
But it was the major presence of the Zeus payload among enterprises that was most surprising to Damballa: The Trojan is known mainly for stealing users' online banking credentials and other financial data. "We've seen Zeus a lot" in enterprises, Ollmann says. With all of the DIY toolkit plug-ins for the Trojan, it's more "versatile" for targeted attacks, he says. "If you want to target a particular organization or vertical, there are lots of plug-ins or automated tool sets to optimize data extraction," he says.
During one six-week period last year, the so-called ZeusBotnet had infiltrated hundreds of thousands of enterprise machines. "It just exploded over that period and nestled itself inside the enterprise. Then it died overnight," Ollmann says. He says the criminal gang behind this botnet at its peak had control of more than 600,000 enterprise bots and was able to get into several million other machines.
But Damballa's list covers only the top 10 botnets that contain named malware for which there is an antivirus signature. "Over 60 percent of the malware we see associated with botnets has no AV signatures," Ollmann says. "The real damage to enterprises is what's outside the Top 10. But they still haven't managed to overcome the big, noisy, [and known] threats that account for 80 percent of all detections."
When an enterprise becomes aware of a botnet infection, it often first goes after the infection hitting the majority of the machines. "They see one botnet accounts for 19,000 of their infected machines, and another, 200. Most will focus on the big one, but the other one is probably a smaller botnet targeting that organization looking for specific information about that organization," Ollmann says.
A typical botnet built to recruit enterprise machines is about 1,000-strong, while a big-name spamming botnet can be anywhere from 50,000 to hundreds of thousands of machines, according to Damballa's data. And the average number of botnets found in an enterprise has remained relatively steady during the past couple of years, with 5 to 7 percent of all corporate systems infected by bots, according to Damballa's data. And these are only the machines that are successfully being controlled by the botnet -- many other machines can be infected with the bot malware, but just not be under the remote control of the botnet. "Within a large enterprise of more than 12,000 nodes, you would expect to see two dozen different botnets running and operating at any particular time," Ollmann says.
Behind ZuesBotnet, KoobfaceBotnetB, ClickfraudBotnet, and SpamfraudBotnet on the Top 10 list were MonkifBotnetA (8 percent), KoobfaceBotnetD (5 percent), TidservBotnet (5 percent), MonkifBotnetB (4 percent), KoobfaceBotnetC (4 percent), and ConfickerBotnetA (4 percent).
Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message.