More Researchers Going On The Offensive To Kill Botnets

Yet another botnet has been shut down as of today as researchers joined forces with ISPs to cut communications to the prolific Lethic spamming botnet -- a development that illustrates how botnet hunters increasingly are going on the offensive to stop cybercriminals, mainly by disrupting their valuable bot infrastructures.

For the most part researchers monitor and study botnets with honeypots and other more passive methods. Then security vendors come up with malware signatures to help their customers scan for these threats. But some researchers are turning up the heat on the bad guys' botnet infrastructures by taking the lead in killing some botnets: Aside from last weekend's takedown by Neustar of Lethic, which is responsible for about 10 percent of all spam, FireEye last November helped shut down the MegaD botnet. And researchers at the University of California at Santa Barbara in May revealed they had taken the offensive strategy one step further by infiltrating the Torpig botnet, a bold and controversial move that stirred debate about just how far researchers should go to disrupt a botnet.

Back in 2008 after two major ISPs halted traffic to malicious hosting provider McColo, spam worldwide dropped around 70 percent because McColo had been the main home to most botnet command and control (C&C) servers.

But deploying more offensive tactics to stop botnets and bad guys is not so straightforward: Researchers walk a fine line as to how far they can go legally and ethically, and sometimes taking down a botnet actually backfires, either with the bad guys returning the favor with a denial-of-service (DoS) attack, or learning how to better evade investigators next time. There's the danger that getting inside a botnet will just give its operators more tools and insight into how to strengthen their operations; botnet operators are notorious for reinventing themselves with stealthier botnets and new forms of malware.

Still, some researchers say playing so much defense and relying on user education today just isn't working in the fight against cybercrime.

"Preaching to people to improve their security only goes so far. It is time to take the fight to cybercriminals themselves," says Marc Maiffret, chief security architect for FireEye. Maiffret says this means getting out ahead of the botnet, like FireEye did with MegaD, or exploring ways to disrupt these operations.

But that doesn't mean manipulating or interacting with infected bot machines, he says. "You have to proceed with caution. You never want to do anything against infected computers. But if you have a way of getting those computers by way of normal propagation come back to your sinkhole, you can fairly easily disrupt" a botnet like FireEye did with MegaD, he says. FireEye was able to "blackhole" the communications to the botnet.

"FireEye chopped the head off the thing," Maiffret says. "That's one way of going after these guys."

But while taking down a botnet has obvious short-term benefits and chalks one up for the good guys, experts say it can also backfire. Gadi Evron, an independent security strategist based in Israel, says there needs to be a more offensive approach to taking down cybercriminals and botnets. "But that approach still needs to be defined [and collaborative]," Evron says. That will take a careful and cohesive strategy that likely entails legal changes and ensures any efforts don't backfire and instigate the bad guys, according to Evron.

"We need to careful that we're not starting a war on the Net that we can't win," Evron says."We have a lot to lose in this war -- [the cybercriminals] can just take down the Net. Many of these criminals are heavily invested, but have shown they are willing to play [with] a scorched-earth policy."

Steven Adair, a researcher with the Shadowserver Foundation, says other botnets are quietly being taken down around the world. "But not all get written about," Adair says. Adair says more researchers are hunting botnets today than ever before, which may account, in part, for the increase in botnet busts and debate about how to kill these malicious infrastructures.

And in some cases, he says, taking down or shutting down parts of a botnet just results in the bad guys retrenching elsewhere. "There have been a few cases where ... it has been a wasted effort," he says.

But Adair says the security community should be more proactive. "At some point you need to get involved and do a takedown with the registrars," he says.

Meanwhile, it's unclear whether the recent silencing of the Lethic botnet is a short-term victory or will have long-lasting effects on the spam operation. Neustar worked with ISPs that hosted the botnet's F servers to knock the botnet out of commission. But while the spam was nearly gone as of yesterday, some Lethic bots were still trying to communicate with a new C&C server, according to researchers at M86 Labs who study the botnet. So it was unclear whether some residual pieces of Lethic are still alive out there.

Lethic mostly focused on pharmaceutical, replica, and diploma spam campaigns. When UC-Santa Barbara revealed how it wrested control of the infamous Torpig/Sinowal/Anserin botnet for 10 days last year, the research set off a firestorm within the security community about whether the researchers had gone too far by taking control of the botnet's C&C server. The researchers basically turned the tables on Torpig's own domain-flux architecture, which uses a single domain name that maps to multiple IP addresses that rotate in a round-robin fashion to evade detection. They registered the domain names before the infected machines were able to be programmed to contact the C&C servers, so the researchers were able to see firsthand what kind of financial data Torpig was gathering and which machines were getting infected.

"A lot of this malware is passively sitting there and waiting for you to do certain actions, so researchers [typically] have to trigger those actions or reverse-engineer the malware," says Brett Stone-Gross, one of the UCSB researchers. "But if you're sitting in it ... you can see exactly" what's going on, like UCSB did, he says.

Stone-Goss says the researchers were concerned about the legal ramifications of getting so entrenched into the Torpig botnet. They contacted the DoD and FBI to make sure they didn't mistake them for malicious activity, he says. "They were actually excited about what we were doing," Stone-Goss adds.

UCSB did not send botnet commands to the infected bots, he says. "But we could see what information was coming [into the botnet]" -- mostly banking information, FTP credentials, and email addresses were stolen by Torpig, he says.

Stone-Goss maintains that shutting down a botnet is effective because it costs the botnet operators so much: "It makes it more expensive for them to move around and they lose data," he says.

Still, researchers agree it's getting harder and harder to get to the core of the botnet problem: the perpetrators themselves. Many of them run like a typical enterprise, with suppliers and partners. Gunter Ollmann, vice president of research for Damballa, says not only are these bad guys well-insulated from the botnet infrastructure by multiple layers, but their operations are especially complex and sophisticated. "It's getting a lot harder to get to the guys behind this," Ollmann says. "Some botnet operators sell botnet agents to other operators. It then becomes more difficult to get to who's in control."

But researchers say they'll keep trying to go after botnets and shutting them down. "We are not actively going after any others right now, but I am sure when the timing and botnet is right, we will again," FireEye's Maiffret says.

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.