Reports that security researchers are running scared from hackers responsible for the Storm Trojan are overblown, say some of the people who have dug into the complex malware.
Earlier this week, Josh Corman of IBM's Internet Security Systems said that Storm, a multifaceted Trojan Horse that has been used to gather a substantial army of bots (or compromised computers), strikes back using distributed denial-of-service (DDoS) attacks when it senses probes of its command-and-control network. These attacks, Corman said, have researchers spooked.
"They're afraid," Corman told attendees at Interop New York. "They find these things but never say anything about them."
Not even close, said two researchers.
"No, we're not scared," said Joe Stewart, a senior security researcher at SecureWorks who has been rooting around Storm since it first appeared earlier this year. "Cautious, possibly. We're still going to do our research."
Zulfikar Ramzan, a senior principal researcher at Symantec, another vendor that has been posting results from its Storm research for months, agreed. "I don't think it's made [researchers] more scared. They're still publishing."
Corman nailed the attacks, however, Stewart and Ramzan said. Both confirmed that they knew Storm had launched DDoS attacks, and as Corman pointed out, that the Trojan has an automated early warning system that sniffs probes made of the botnet. "Storm understands any attempt to understand it, then notifies the bot controller," said Ramzan. "It seems to recognize a threat after several different attempts to probe the bot."
The tactic isn't new, but Storm has taken it to higher levels of automation, said Ramzan.
Attacks against spam and malware fighters aren't anything out of the ordinary. "This has happened before to different groups that seem to threaten the criminals," said Stewart. "Spamhaus undergoes attacks fairly frequently, for example."
There are ways for researchers to disguise themselves or cover their tracks if they're concerned about hacker counterattacks, both researchers said, although using them can be dicey for those on the right side of the law. "You have to be careful," noted Stewart. "You don't want to have some innocent ISP to end up the brunt of an attack. We don't want any collateral damage."
Symantec examines bots built by Storm, and other such malware, in a controlled, isolated environment, added Ramzan. The emulated network lets researchers ping and ding the bot without any chance that it will "talk" to the outside world, and so reach the botnet controller.
As for why Storm retaliates, each researcher had an explanation. "Hackers discovered that they can get away with DDoS attacks," said Stewart, using a variation of George Mallory's famous "because it's there" rationale. "The really big one was on Blue Security last year. No one was ever charged, and Blue Security had to take down its business. That was a win-win for the hackers, so why not go after others?"
Blue Security, an Israeli-based antispam company, threw in the towel in May 2006 after days of crippling DDoS attacks launched against its network and Internet provider. Blue Security was best known for its community-based "Do Not Intrude" program that spammed spammers with large numbers of opt-out requests from the half-million users who had signed on to the concept.
"The only thing I can think of is that they're trying to send a message of some sort," said Ramzan, who pointed out that by its nature, a DDoS attack exposes the botnet itself to retaliation or take-down. That's why DDoS-for-pay has been decreasing, he said. "It's a dangerous way for a bot master to make money, because they're exposing their entire army of bots to you. Why they would do that to hit back, I don't know. Maybe they think they have bots to spare."