New Phishing Exploits Emerge

The phish are biting, alright -- and your users could be next on the menu.

The Anti-Phishing Working Group (APWG) yesterday released its latest phishing trends report, which shows that the number of different "brands" (organizations) reporting phishing attacks jumped from 97 in April to 137 in May: the most since December, when it was at 121. "That's a pretty severe" increase, says Peter Cassidy, secretary general of the APWG and director of research for Triache, a consultancy. "It had started to look like conventional phishing had leveled off."

There were more newly detected unique phishing Websites last month, too, according to the APWG report, up to 11,976. Not surprisingly, financial services organizations are still the main victims of these attacks, at 92 percent.

While phishing is spreading, however, many observers are more worried about the improved quality of attacks than about their quantity. Experts who track the phishing scene say they are seeing a new wave of exploits that go far beyond the old Website-hijacking scams.

Keyloggers are among the most sophisticated and fastest-growing types of phishing attacks on the Web, says Dan Hubbard, head of Websense's Security Labs and member of the APWG. The keylogger's goal: Infect user machines and, ultimately, steal their data and privileges.

"There's a shift in using malicious code and exploits to infect users instead of Websites with a simply deceptive tactic," Hubbard says.

Phishers are becoming more efficient, too, often reusing a single keylogger to attack multiple sites. Hubbard has seen at least one exploit in which a phisher used the same keylogger to attack more than 1,000 banks.

Host-file redirector phishing attacks, which change files in user machines and redirect them to nefarious Websites, are a close second to keyloggers, Hubbard says.

Phishers are also finding creative ways to avoid getting caught. They're now recycling and reinventing some of the same malware code so it's tougher for law enforcement and antivirus tools to detect. "So you see different iterations of the same crimeware over and over, and it's harder for [antivirus software] to detect," Cassidy says.

All it takes to evade an antivirus scan is changing one or two bits of information, Hubbard observes. "A lot of phishing schemes are refactored code."

At the same time, phishing toolkits are getting easier to obtain. The most infamous toolkits you can buy off the Net are WebAttacker and Nuclear Grabber. WebAttacker lets a PC novice implant exploit code on his or her Website so that visitors will get infected when they surf the site. Nuclear Grabber lets an attacker sit on a real banking site and grab data from electronic forms. Both of these tools are hosted on Russian Websites, and they aren't cheap: Nuclear Grabber, for instance, costs $3,000.

As the availability of phishing toolkits grows, it is becoming more difficult for law enforcement to track the attackers. Cassidy says phishing gangs sometimes give away toolkits in order to distract investigators from their own operations. "Then they'll get amateurs involved to obscure their activity."

And here's an eerie look at the shape of things to come: The Brazilian Computer Emergency Response Team (CERT) recently found what security experts say is the first "phishing worm." The worm, which hit CERT Brazil's "honeypot" systems, installed crimeware akin to what's used in a phishing attack and behaved similarly, watching the victim's Web activity and mimicking legitimate Websites, according to Cristine Hoepers, general manager of CERT Brazil. And it sent all the data it collected via email.

The worm, dubbed Net-Worm.Win32.Banker.a or W32/Banker-BIX, only works with the Brazilian Portuguese version of Windows, and AV vendors have since released updates for it. But it sent a chill down the spine of security researchers because the victim doesn't have to open a file or click on a link to be infected. "It's a monster," says APWG's Cassidy. "It doesn't require you to execute a host…you don't have to open anything. It just bores its way through networks and plants itself on machines."

Another worrisome exploit on the horizon is the "man-in-the-middle" attack. Rachna Dhamija, postdoctoral fellow at the Harvard Center for Research on Computation and Society, says it's possible for phishing-type attacks to hit PassMark -- the two-factor authentication scheme used by banks such as Bank of America, which requires users to see an image they had previously selected before they enter their password. This authentication scheme is proving to be vulnerable to an "active" man-in-the-middle attack, where the attacker is truly a person sitting on the server and capturing the user's password and credentials. "We are on alert for this" type of attack, she says.

Freeware tools such as Ettercap can also be misused to launch a man-in-the-middle attack, says Sean Kelly, business technology consultant for Consilium1. Such attacks would be tough to detect. "A phisher could set up and pose as a Website."

In the end, though, the phishing problem is still a user interface problem. Email and browser user interfaces are just too easily spoofed and manipulated, Dhamija says, and security professionals and vendors should expect more attacks from that angle. "Phishers are really good at designing user interfaces."

— Kelly Jackson Higgins, Senior Editor, Dark Reading

Companies mentioned in this article: