Verified by Visa and MasterCard SecureCode

Steven Murdoch and Ross Anderson raise many good points, including the phishability of 3D Secure, and the way it will train people to stray from standard security advice.

One quibble: I wouldn't call 3D Secure a "Single Sign On" system. If I have two Visa cards from different issuers, then my 3D Secure experiences will not interoperate, and will look and feel quite different, because as the authors point out, the real time authentication mechanism is left to the choice of the issuer.

They're right of course that economics trumps security in this exercise.  Yet there's still a stark architectural question that begs answering in all this.  What really necessitated such a radical change to the decades old Four Party card settlement model? 

An important feature of the Four Party model is that it means customers’ transactions are distanced from the issuer; as the authors say, this is good for customer privacy.  It’s also architecturally elegant for payment information to flow between cardholder and merchant in real time, and between acquiring bank and issuing bank in batch time.

In contrast, 3D Secure joins the cardholder to their issuer in real time, to effect the authentication step. This forces unusual and inconsistent behaviour onto the user, creates bottlenecks, and probably creates new vulnerabilities due to the sheer complexity and novelty. 

K.I.S.S.!

Hi Stephen,

Thanks for your comment. I think the architectural change to the four-party model which 3DS introduces is a useful insight. Your point about whether 3DS is single sign-on is also interesting, so I have published a separate blog post on this question.

Steven.

Sorry, the URL to my comment was wrong. It should be http://www.lightbluetouchpaper.org/2010/01/29/why-is-3-d-secure-a-single-sign-on-system/.

What is most telling about 3D Secure is that it merely adds a layer of complexity rather than get to the root of the problem in the CNP space ' "is John Smith using John Smith's payment card?".

People point to the low fraud numbers with VbV and MSC as proof positive the system works, the fact is that as only 200 million people (perhaps less) use the system the feeding grounds for hackers is too small for them to start poaching. 

PayPal went from being the most secure payment system to the most hacked as its numbers grew, making it a more attractive target for hackers who need a lot of activity to mask their work. The same will happen to 3D Secure.

The holy grail is for a CNP anti-fraud system to offer the same level of protection as a customer walking itno a shop, presenting his payment card and proof of identity before the merchant gives him his purchases.

In the interests of full disclosure, my company GenMobi Technologies will be unveiling later this year a new technology that will offer a more secure verification that "John Smith using John Smith's payment card" without the need for the consumer to memorize PINs or passwords.

Just to clarify, there are a lot of people who read Finextra, like me, from the US.  Neither of these programs have gained serious adoption here.  The banks stopped promoting it actively years ago.  Something like ~5% are enrolled but that includes people like me who did it once and now avoid it like a root canal.  Just the other day, I was buying my daughter a plane ticket on Jet Blue and when I noticed they supported VbV, I simply switched and paid with my Discover Card.  It was easier to do that than remember what my password from 4 years ago might be for my Visa card.

There has been a lot of talk in the last 2-3 years of updating the program to default to risk based authentication, mostly behind the scenes, rather than interrupting the check-out experience.  Some issuing banks are now supporting this approach whereby the consumer, enrolled or not, would only see the authentication frame if some attribute of the tx looked suspicious like cardholder from US and IP from Singapore.  This would work like many of the 3rd party fraud screening solutions work.