With Web 2.0, a new breed of malware evolves

Web 2.0 technologies may be laying the groundwork for a new generation of hacker tools, a noted security researcher said Wednesday.

Google Mashups, RSS feeds, search, all of these can be misused by hackers to distribute malware, attack Web surfers and communicate with botnets, said Petko Petkov, a security researcher speaking at the Open Web Application Security Project (OWASP) U.S. 2007 conference, held on eBay's campus.

Tools like the downloadable MPack hacker toolkit have made it easier for the bad guys to deploy malicious code, but some of these emerging technologies promise to take hacking to a whole new level, he said. "Now people can use and abuse Web 2.0 technologies to construct something much larger," he said. "When you look at it from a hacker perspective, you'll see there are a whole lot of opportunities," he said.

For example, it took Petkov just one day to build a Web-based attack infrastructure using Google Mashup Editor, Google's invite-only Web application development service. And even if Google decided to shut down this type of attack service, its open and distributed design makes it very easy to set up a new account and launch an identical service. "What is cool and what is the problem is if somebody closes the application down, these attackers can bring back the application to a live version in five seconds," he said.

This kind of Web 2.0 malware is in its infancy, but it's starting to be used, said Wayne Huang, CEO of Web security firm Armorize, based in Santa Clara, California. Huang says he has seen attackers use Google alerts to scan the Web for sites that are running software with known vulnerabilities, and he said that criminals are also starting to use RSS-to-e-mail conversion services to have an untraceable way of controlling their networks of hacked computers, called botnets.

Until recently botnets would always look for commands on a pre-allocated IRC channel, but now, distributed RSS-based command-and-control networks are coming into favor, Huang said. This makes it much harder for law enforcement to take down the computers that are actually sending the instructions to the botnet machines.

Another tactic: Attackers could encode the date and domain name of the computer looking for instructions on random Web pages that would also contain attack instructions for the botnet. These instructions could then be dug up by the botnet using Google search.

Researchers like Petkov and Huang believe that criminals are only beginning to experiment with Web 2.0 hacking techniques like these, but that if they do catch on, it could become a nightmare for the Web 2.0 world.

"Nobody realizes the potential for abuse," Huang said. "When it happens, I think it's going to be on a very massive scale and very hard to stop."

"I think [these attacks] are brand new right now, but with time they're going to get more relevant," Petkov said. "Right now we're still in the playground with this."