PhishTank, a new service from Internet-acceleration startup OpenDNS, wants to rid the Web of "phishing" emails - messages that pose as legitimate customer-service emails from banks and e-commerce websites and try to trick users into revealing their passwords. The site works by collecting examples of suspected phishing emails, and then relies on users to vote on them as legitimate or illegitimate.

The digerati are loving it: Georgia Tech student and former Yahoo intern Paul Stamatiou praises PhishTank for a simple phish-submission process and an open API, or application programming interface, which lets other websites and software makers tap into PhishTank's database of suspicious emails. "Once the PhishTank databases grows, other sites can harness the data using open APIs which will remain free," writes Digital Inspiration blogger Amit Agarwal.

We hate to rain on the PhishTank parade, but that's a mighty big if. What's the incentive for ordinary users to report phishing emails? They're busy enough just deleting spam - and for mere mortals, the distinction between phishing email and spam is exceedingly fine. Our suggestion: Rather than relying on consumers to take time to sort through their emails and report suspected phishes - let alone vote on them after they're submitted - PhishTank should focus on making alliances with big email providers like Microsoft, Google, and Yahoo, to act as a central clearinghouse for anti-"phishing" data.