It appears that many experts are on the side of recommending against e-commerce altogether. Just too dangerous. I was recently involved in a discussion on the famous Funsec security list with several security experts who argued the same thing about online banking.
I would argue that the benefits of online banking, and e-commerce for the most part, are so huge that forgoing them is an overreaction. The right way to approach the security issues with online banking is to be aware of them. This is, sad to say, a lot to ask of the average Dick and Jane on the Internet.
As is clear from the Funsec discussion, the problem is basically one of e-mails. While there are cutting-edge moves into other media, the overwhelming majority of phishing attacks are initiated through a fake e-mail. The experts in the discussion were upset at legitimate e-mails from banks that bore serious resemblance to phishing attacks.
Another example reported by security vendor Sophos dealt with an e-mail from Citibank Australia that followed the absolutely classic phishing script of asking the user to click on a link in the message and follow it to the banks Web site, to authenticate the users card number, account number and PIN number.
This is a serious lame-brained example, and someones career should suffer for sending the e-mail. But this is atypical, and some banks are better about this sort of thing than others.
The idea gaining currency that banks should never send e-mail to their customers strikes me as a dangerous one. Im a Bank of America customer, and I use online banking extensively. I very rarely get e-mails from the bank, and most of them, such as a notice of direct deposit, are things I specifically asked to be sent. There are few links in these e-mails and none that say, as in a phish, "Click this link to solve all your problems."
And anyway, even if banks dont send out any e-mails, it wouldnt solve the basic problem that users need to know that they shouldnt expect any e-mails. If users were smart enough to know that any e-mail that seemed to be from the bank was illegitimate, they wouldnt get scammed anyway. Once your suspicions are aroused its just a small step to seeing that any phish is a fraud.
Banks have gotten very good at authenticating themselves to users—in other words, proving to users that the site they are looking at is actually the bank they intended to contact. Its a very different thing for users to determine with confidence that the fake site they are looking at is a fake.
Take my own Bank of America. Because of its SiteKey feature I know Im logging into the real Bank of America site when I log in.
But what about the Bank of America phishing site? It doesnt have SiteKey, but neither does it say "No SiteKey." I have to notice that the SiteKey isnt there before I type in my username and password. Same with other forms of strong authentication.
So what can be done? Part of the answer lies in an emerging generation of anti-phishing tools, such as those in Internet Explorer 7, Firefox 2.0 and Symantecs latest generation of products. These tools all let a lot of phish get away now, but I expect them to get better over time.
The only alternative is to avoid online commerce altogether. One might as well avoid anything that has a potential for danger. The answer is not to cave in, but to educate.
Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983.
More from Larry Seltzer