Can a small startup in San Francisco play a big role in the fight against the phishing scourge?
OpenDNS, a 10-employee company with technological expertise in the domain-name-resolution field, is betting that a mix of social networking, Web services and user-generated submissions can power a valuable, freely available anti-phishing database.
Phishing, which has emerged as a billion-dollar underground scam, is the practice of using well-disguised fake sites to trick users into entering sensitive information such as credit card numbers, passwords, addresses and Social Security numbers.
With the new PhishTank, OpenDNS has set up an easy way for anyone to submit suspicious fake Web sites to a repository and release the data via a free API to let programmers build anti-phishing tools into their own programs and Web applications.
"This is the opposite of the conventional approach where the Symantecs and McAfees collect the phishing data and never share it," said David Ulevitch, CEO at OpenDNS, in an interview with eWEEK. "When you toss [a phishing site] into the PhishTank, they all have access to it. Anyone can use the data freely to build security applications."
Borrowing from the Digg model of using the power of a motivated community to drive submissions, OpenDNS has set up a voting system that lets registered users vote on the validity of a site dumped into the PhishTank.
This process removes the tedious task of manually verifying submissions and ensures that only legitimate phishing sites are put into the database.
OpenDNS provides an easy-to-use system for bypassing domain-name-resolution services offered by ISPs, promising to speed up Web page loads and protect surfers from landing on dangerous sites. In exchange, the company serves advertising on landing pages if a user chooses to have the technology correct mistyped domains.
Ulevitch said OpenDNS will be using data from the PhishTank to improve its service and stressed that it was in his company's interest to make the clearinghouse "as big and as open as possible."
In the first iteration of the service, users can submit suspicious sites via e-mail or directly on the PhishTank Web site, but once the data bank gets filled and developers make use of the API, Ulevitch said he expects to see useful applications and add-ons emerge from the community.
For advice on how to secure your network and applications, as well as the latest security news, visit Ziff Davis Internet's Security IT Hub.
"It's been live for a few days, and we've already seen [a Microsoft] Outlook plug-in. Someone else is working on a Firefox extension. That's the kind of community element that is needed to make this truly open and useful for everyone," said Ulevitch. "I can think of many ways to use this data. It can be built into SpamAssassin or any piece of anti-phishing technology that needs verified data."
Any tool or application built with PhishTank data will be featured on the Web site, and Ulevitch plans to actively encourage and endorse those products.
On the back end, OpenDNS automatically creates screen shots of the submitted phishing site and a simplified interface for users to vote on the accuracy of the submission. "Every [submission] requires a threshold to pass before it becomes a verified phish. Once the community verifies it, we know for sure it's a legitimate phish, and that gets put into the database. That info is then fed out to anyone using the API. It's a very accurate way to test these sites," Ulevitch said.
On Oct. 5, a total of 775 suspect sites were submitted to the PhishTank. More than half (462) were verified as legitimate phishing sites.
The site keeps track of the top submitters and top verifiers as a way to reward the community.
Registered users also get a personalized RSS feed to allow them to keep track of submissions and verifications.
OpenDNS isn't the only company using the wider community to tag malicious phishing sites. Sunbelt Software has partnered with CastleCops to launch PIRT (Phishing Incident Reporting and Termination), a community dedicated to taking down phishing sites. PIRT relies heavily on volunteers to submit phishing scams for takedowns. Once a phish is confirmed, the group generates reports and e-mails that go out to various security research outfits.
Secure Computing, in San Jose, Calif., also offers a free online notification service that allows legitimate organizations to register for notifications of online fraud attempts.
The service, at PhishRegistry.org, uses a proprietary technique called Phisherprinting to determine a probable match of any two Web sites via automation.
Out of the Net, Into the Tank
OpenDNS is using a mix of social networking, Web services and user-generated submissions to populate its anti-phishing database:
Users vote on the validity of submitted phishing sites.
An open API lets developers build anti-phishing tools into their own programs and Web apps at no cost.
Personalized RSS feeds keep the community in the loop on verified phishing sites.
Check out eWEEK.com's for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzer's Weblog.
Channel News and Analysis 
