[Cryptography] cost-watch - the cost of the Target breach

[Jerry Leichter]

On Dec 6, 2014, at 2:29 AM, Henry Baker <hbaker1 at pipeline.com> wrote:
> At 07:49 PM 12/5/2014, Jerry Leichter wrote:
> "It's also not at all clear that the banks were the ones who resisted on chip and pin.  *They* wouldn't be the ones bearing the costs of replacing all the card readers out there - and they stand to gain from the liability shift that leaves merchants who don't get new terminals stuck with any loses.  Over all, win/win for the banks."
> ---
> Ross Anderson has been analyzing chip&pin for years & found that there are just as many problems with chip&pin as with the magstripe cards.
I think that's a gross overstatement of Anderson's finding.  Problems, sure.  But attacking chip and pin requires a much higher level of sophistication than attacking magstripe cards.  As George Orwell said, the best is the enemy of the good.  It's not as if there's been a realistic, significantly better, alternative out there.  (Perhaps Apple Pay and similar token-based systems will prove to be the next step.  But new systems *always* look much stronger - until people start attacking them.  We'll see.)

> Ross points out (if I recall his comments correctly) that with chip&pin, the burden of proof moves away from the banks, which is why the banks are so hot for chip&pin.
There's a difference between the British and US legal systems - and always has been.  For years, British banks were successfully able to push liability on to customers (they had their own customers prosecuted for fraud for claiming that withdrawals made using the bank's "completely secure" ATM cards weren't theirs) and merchants.  In the US, they've never been able to push anything on to customers, and in the case of merchants, it's a messy system (though the merchants do usually get stuck).  With chip and pin (but more on that in a moment) they seem to have deployed both carrot and stick:  Once it's in place, any merchants who aren't using it will get stuck with the entire loss, while they have some degree of protection if they are using it.  However ... just to be clear, most banks in the US aren't even rolling out true chip and pin.  They are going with a bizarre blend called chip and signature.  This is better in one way:  Cloning a chip card is, the last I heard, impossible (or at least very expensive), while cloning a magstripe card is trivial.  But you don't get the true second factor of a pin since no one actually checks signatures.

>  But don't hold your breath waiting for chip&pin to produce any improvement.
I don't know where you get this from:  It's produced significant improvements where it's been deployed.  Crime based on cloned cards - the major kind of CC crime in the US - disappeared and hasn't returned.  New, more sophisticated attacks (e.g., based on modifying terminals - sometimes way up the supply chain) have emerged and overall levels of attack are rising.  Of course, the US, coming late to the party, won't have the honeymoon period as the attacks are already out there.

>  I just read that the new US chip&pin system has already been hacked, and it isn't even in real service here yet!
Reference?  (And again, keep in mind that the US is mainly rolling out chip and signature.  The goal is to get to chip and pin eventually, but the industry claims that going there in one step is too much change for consumers to absorb in one shot.)
                                                        -- Jerry