After scanning 2 million computers over the past 24 hours, IBM’s Internet Security Systems (ISS) division said Thursday that it had spotted Conficker on 4 per cent of the IP addresses it monitored. Experts had pegged Conficker infections in the 2 million to 4 million range, but IBM’s numbers suggest that they may be much higher than that, perhaps in the tens of millions.
Although Conficker is clearly the worst worm outbreak in years, the results came as a surprise, according to Holly Stewart, a threat response manager with ISS. “It is higher than what we expected; I thought we’d see 1 to 2 per cent,” Stewart said.
Late last week, IBM researchers reverse-engineered Conficker and figured out a way to track infections by measuring peer-to-peer traffic on the network. They used that technique to reach their estimate.
The results are similar to numbers released Wednesday by OpenDNS, which said it had also spotted a much larger number of infections than expected. Both IBM and OpenDNS’ numbers count Conficker.C, the latest variant of the worm, and one that is easier to spot communicating on the network.
Conficker began spreading in October 2008, using a handful of sneaky tricks to spread. Once it infects a machine, it can spread very quickly on a local area network by taking advantage of a now-patched flaw in Microsoft Windows.
Still, Stewart cautioned against concluding that 4 percent of Internet users had been infected. “It’s not a perfect number, nothing is. But it’s the best that we can give with the data we have right now.”
It’s possible that Conficker infections are approaching 4 percent, said Danny McPherson, chief security officer with Arbor Networks. Because Conficker is more likely to infect certain types of users — broadband consumers are generally more vulnerable than enterprise or government users, for example — estimates like ISS’ could come from a sample that does not represent the Internet as a whole, he said.
Still, by any measure, Conficker is a big problem. “Even if they’re off by an order of magnitude — which is possible — the number of infected machines is immense.”