I was recently asked to comment on the new Chip & Pin attack created by Prof Ross Anderson from Cambridge University. In my original comment released to the press I make an assertion in relation to a change in process that “breaks the circuit” of this attack – see below:
Jay Abbott, director in charge of Threat & Vulnerability Management, PricewaterhouseCoopers LLP (PwC), said:“Essentially, what the scientists have come up with is a very effective and simple way of exploiting weaknesses in the system.
However, it is important to bear in mind that the fraud requires a very specific scenario to become effective.
“A simple process change by the retailer of asking for the card holder to hand over the card would break the circuit, although this isn’t always possible as sometimes the card reader is fixed to a point on the other side of the counter.
“At present, the customer is accountable for the fraud as banks argue that PIN verified transactions are secure. Given this attack demonstrates a clear method of bypassing the PIN system, this assertion by the banks stands on shakier ground.”
With the original comment came a caveat, which as you would normally expect, was not quoted by the media, this caveat was that the process change suggested brought with it the opportunity for cards to be skimmed, which was in fact one of the original reasons behind the Chip & Pin changes. In fact, the change works in the favour of the retailer rather than the consumer, however, before you hang me, allow me to demonstrate the rationale behind this.
Consider first that Chip & Pin is in fact “two factor” authentication, which anyone in the security business will explain is more secure than “one factor” authentication. The first factor is the card itself or the “chip” in this instance, the second factor is the “Pin” which in this context operates as a pass code. Given both elements are authenticators in their own right, both are required, and as such any attack must include them both. The attack designed by Prof Ross Anderson targets the Pin aspect of the authentication, and relies on the original card accessed through a series of technology components that have to be connected together in some way. The method shown in this attack makes use of concealment to hide these components on the person of the attacker, and relies on a custom built “attack” card with wires hidden up the sleeve of the attacker, back to the other components involved. The obvious way to therefore detect and prevent this attack at the retailer is by separating the card from the attacker, thus showing the wires and revealing the ruse.
The cloning of cards must be treated separately as the current methods of cloning (that I am aware of at this point in time) only create “yes cards” which would not work in this attack scenario as they are not true copies and would be detected by the PoS equipment as fraudulent. As I understand it, there is no economically viable way of cloning Chip & PIN Cards effectively at this time. Any cloning would still focus on the magnetic stripe data, which can be easily cloned, but is not accepted by the retailers (usually) when a Chip & PIN card is presented. This of course is at the discretion of the retailer and out of the control of the consumer or the banks.
This brings us to the counter argument, specifically in relation to the increased risk of your card getting skimmed/cloned by the retailer when you hand it over. Een if it were viable to clone the chip cards, given that a card skimmed by a retailer would typically not get the pin as well (this of course is not always the case), using the now cloned card would have to make use of Prof Ross Anderson’s attack method, which if the aforementioned process change was implemented, would not work, so in effect increasing the risk of cloning, but decreasing the risk of a successful attack using the cloned card and “breaking the circuit”.
This of course relies on the premise that the use of the cards magnetic strip is in fact not viable, and therefore if anything, reinforces the use of Chip & PIN ironically. Of course in real life the Magstrip is regularly used, but that, again is outside the scope of this discussion and considered irrelevant in the face of the specific discussion around Prof Andersons attack.
There is always of course the argument for using a small form factor wireless transmission device to remove the need for wires, but given the form factor of a credit card and the inability to alter this form factor without raising suspicion, I am personally unsure that significant enough range for a TX/RX comms loop could be achieved given the power that could be implemented into a credit card sized device.
Again, in my original comments to the press I clearly stated that the system needed to be fixed, and that the attack was effective, so this is not me suggesting that we should brush this under the carpet, in fact it is simply looking at what we can potentially do NOW to protect the system, while its eventual upgrade is debated and planned.
Don’t forget, in this context I am just as much of a concerned consumer as you.