Hacker News new | past | comments | ask | show | jobs | submit login
When Undercover Credit Card Buys Go Bad (krebsonsecurity.com)
161 points by kawera on Dec 8, 2015 | hide | past | favorite | 80 comments

I propose that a few fake credit card numbers and names be issued to different companies and that they place them in their records. If there is a breach, finding those fakes will identify where the breach came from. Each set of fakes would only be given to a specific company.

This is similar to how dictionaries (before the internet) would post fake entries to discover if competitors were copying them. Or how (some claim) map makers would plant fake places to identify similar copying. In the credit card case though, it wouldn't harm innocent users who may mistakenly think the fake is real.

That wouldn't work.

Your assumption is that these retail companies are storing full credit card information in some large database somewhere, that is not the case at all. Most large retailers only store the full credit card information for seconds while they process the transaction, when the transaction is complete it is discarded.

What the criminals are targeting is the card terminals themselves. Meaning when you swipe your card, as the transaction is being conducted they take a copy of the information and re-transmit it to a host somewhere they control (or to a C&C machine on that network, which they then exfiltrate somehow).

Plus even if you could get your "gotcha" credit cards into the criminal's hands, the chance of law enforcement buying one back is tiny, since the stolen batch is mixed with other cards, and only then a small sub-set is sold (since they verify if the card is still active, and meets other conditions).

My point is, it just won't work with the situation we're dealing with today. Chip & Pin may help reduce retail card theft as it is harder to reproduce a physical card to later use, and internet transactions should require the CSC which a brick and mortar store won't record (even momentarily).

Not downvoting, but the number of corporations storing credit card information would terrify you.

Seen it at former jobs as a dba, helped to change that at some, and its among the top ten list of things to check when I start at new one.

If they got caught doing that, both VISA and Mastercard will essentially ban them. At least if it is an insecure as you make it sound and they haven't met PCI DSS.

So instead of storing known gotcha cards in these databases, a likely better solution would be to discontinue the use of these databases.

Most major retailers don't store credit card numbers. The full information never leaves the terminal, and is discarded when the transaction is complete.

PS - Thanks for "Not downvoting" but why would you anyway? Nothing I said contradicts anything you've said. I said large retailers aren't storing CC information. Are you claiming they do? If so, name names.

There are plenty of small "mom & pop" sites out there that store full credit card info, in plain text, in unsecured databases, in blatant disregard of card company policies. They go with the thinking that it costs to much to implement security and nothing will ever happen to us. We hear about the big breaches on the news, but I would bet for every big breach we hear about on the news there are thousands of smaller breaches that don't get recognized.

For example, about two years ago I had a property management company ask me to build out a site to handle rent payments for them. They wanted to store and have access to the full credit card numbers to manually process transactions for failed or late payments. When I told them they were not allowed to do that they terminated the contract with me. I know they went on to hire a developer to build out a site for them and store credit card numbers. While I don't know if they are still doing it, it did happen for an extensive period of time.

That was also the case for a number of ACH portal softwares sold alongside retail POS suites in the 2000s. Instead of temporary storage until batching, full transaction details were held permanently by default.

At any time, anyone with access to a payment terminal who knew where to look could process transactions manually from these records. They could have double charged and skimmed, done reversals for a cut, or copied and sold customer profiles with CC numbers. These were BSA products marketed to small and medium sized businesses, so no one questioned compliance.

I prefer to put consequences in terms of money.

That said, on the flip side, because of things like "Dual Control" the PCI DSS is physically impossible for some small companies to abide by. In the past we dealt with credit cards and had a few audits, and they admitted that many of the points are aimed at companies that don't consist of just two people.

Also, look at it from their POV. In a normal course of business a lot of these mom and pops probably already are storing credit card information on paper and/or in word documents. So you're saying the system you build for them is going to be, in some ways, less useful than their current ad-hoc system.

My mom & pop tax guy emails me my tax return as an encrypted pdf, with the password being... the last 4 digits of my SSN.

Oh there is no question if you bring paper records in to the mix there are enormous opportunities. I worked at a mom & pop restaurant for several years. In the attic of the restaurant were stacks of boxes with credit card receipts dating back to the mid-80s at least. While the chances of many of these cards being valid now are significantly reduced, there is always that slight chance. I have a credit card from the early 90s that the number has not changed (kind of nice in a way as I can recite the number by memory now).

On the flip side though, to some extent, I would consider these records safer than a website. At least to gain access to these records you would have to physically enter the building, climb in the attic, sift through hundreds of boxes, gather the info you want, and then leave. This would also require you be in the general location of the building and have knowledge of the records being there. Hacking a website database from across the globe is a whole lot easier, so you I would say that their existing paper system is more secure to a degree.

You seem to think VISA and Mastercard have the desire to reduce fraud, they dont. They have a desire to reduce their liability.

PCI DSS is largely a marketing ploy, and way for them to shift liability from banks and the network to the merchants.

As a security standard PCI DSS is terrible

Further storing card numbers is not a violation of PCI-DSS, it would be impossible for them require that as reoccurring billing, auto-pay, etc have a requirement for the PAN to be stored.

Look at Amazon, it stores my card number, in full, so I can checkout with out having to reenter it every time

While I do not want to name names, I have worked for a very large retailer that everyone in the US would know that stored credit card information for quite a long time. They are large enough to do large chunks of their credit card processing themselves, instead of having a third party processor. They even have their own set of cards that they process themselves. It might have changed in the last few years though.

Either way, they did take PCI DSS seriously, and passed audits: dual control requirements for keys and all that. Still, accounting could print reports with many credit card numbers, although it was all logged and relatively well tracked.

I have not worked in a lot of big retailers though, but I have seen others that featured searching for a missing receipt by credit cards number, or refunds to the original CC way past the window where a processor can just reverse a charge. There's also online retailers, that definitely remember your credit card number. So I suspect there are plenty of retailers that at least have historically kept CC numbers, whether being PCI compliant or not is another matter entirely.

I worked at a place ~10 years ago that had full PCI certification, you only have to be compliant if your the one doing the processing. We had loads of 3rd party clients using one of the customer facing applications that's stored the card details and potentially the cv2 numbers in the clear. They then sent them to us to batch process a couple of times a day. If their connection crapped out, they sent them again later. I distinctly remember reprocessing card details of transactions that had happened more than 2 weeks prior. Credit card numbers are incredibly easy to generate too, the check digit routine is well known, and you can just try the other details randomly for small amounts till you hit jackpot.

I haven't read Brian's post though because I'm getting 503 errors atm.

You can't really "try the other details randomly for small amounts till you hit jackpot" - merchants who do that (or if their service allows others to do that) get simply kicked out of the network, pretty much every acquirer will monitor your rejection rates closely.

Storing the a cryptographic hash of the credit card number would allow for searching for receipts by credit card number without storing the actual credit card number.

Credit card numbers aren't nearly long enough to prevent a brute force attack against anything of the nature.

I think it partly depends on how many customers the store has. If it's a mom-and-pop place with a salt on their hash, attackers would get much less "profit" out of their search.

Anywho, just going to engage in some half-assed napkin-math to explore the idea:

Card numbers: 6 digits for institution and 9 for the customer and 1 for checksum. Let's assume merely 20 big issuers "worth attacking". The possibility-space to search is then (20 * 9^10), or ~70 million.

Suppose that--by design--it takes 2 seconds to compute the hash. Covering the whole range would take ~4420 years. That's... not a very comforting margin.

Perhaps you could hash the credit-card along with the year-month in which the purchase occurred, but attackers probably aren't interested in >12 month old cards, so that's only a small 12x slowdown there.

If the point of storing them is to search for credit card matches, I don't see how you could salt them in any useful way.

Beyond that, most things display first four & last four, so you can probably assume they have that too.

> I don't see how you could salt them in any useful way

Yes, I agree. Even two seconds of waiting for your search is a little on the high side, UI-wise.

> Beyond that, most things display first four & last four

My impression is that "only last four" is the most-common.

Are card numbers perfectly random? I mean, excluding the industry/company identifier and the check digit?

Whoops, it seems I kept editing without seeing your reply. Sorry, bad habit of mine. I don't know how random the customer-portion is, but I assume/hope it is. I put some napkin-math in, but AFACT javawizard's right about the basic issue.

It might be a different story if the system required users to also enter in data like the exact customer first/last name.

There's a way to generate subsequent AMEX Card numbers once the card is deactivated[0]


The problem is not that the merchant stores a number, but that all the merchants store the same number. Why not just use OAuth?

~8 years ago, I worked at one of the largest retailers in North America (top 5). Credit card numbers were absolutely stored in logs accessible to me (an on-the-floor employee who processes returns, etc.). It was useful as I could use them to find purchases if someone lost their receipt but had the credit card. It was also incredibly insecure. I believe they've changed it since then, but it would not surprise me at all if other retailers were in the same boat.

Agreed, the downvote should be used for content you find inappropriate or adds nothing to the discussion. A dissenting opinion that is on topic should not be downvoted.

Downvoting to signal disagreement is allowed, and you can't stop it in practical terms anyway, so accepting that it is a thing is pretty useful.

Getting a little offtopic, but regarding your PS: I dont know, I just often see someone assume that someone replying to them (when they are being downvoted by others) assume that the person has negative intent and is part of the downvoting spree.

I would much rather engage in some discussion even if their opinion is honestly held than downvote someone I disagree with. If someone is a shill or not listening to reason on the other hand...

I've worked at shops that don't meet PCI DSS. It's a sticking point, and we WANT to comply, but also don't want to spend the money to comply, and the processor wants us to comply too, but they'd rather have our business than give us the boot for not meeting PCI DSS.

A lot of companies do store full card numbers. Not usually physical retail stores, but for online companies it's not /that/ unusual.

Most do just tokenize with a third party processing company though.

This would be a PCI violation, so if it's common, then someone has some calls to make.

I'm a stickler about enforcing data retention (and non-retention) standards where I work. We never store any credit card information. That would be an outrageous liability.

Storing the PAN is not a "PCI violation". It's not best practice, for sure, and you then have an obligation to secure your systems much more thoroughly, but you can do it. "Payment brand rules allow for the storage of primary account number (PAN), expiration date, cardholder name, and service code." https://www.pcisecuritystandards.org/documents/SAQ_InstrGuid...

(There are other types of data you can NEVER store (also described in the above document) like the CVV or PIN code, which is perhaps what you meant.)

The person I was responding to just said "credit card information". CVV, PIN, and CC # are all verboten.

Sorry, you are wrong about the CC# (PAN) Read what I linked to again.

Oops, you're right.

Swiper skimming might be an issue as well, but didn't some of the largest breaches include stored credit card data? Washington Post says credit and debit card information was included in the Target breach.


I said nothing about swiper skimming.

What I said was:

> What the criminals are targeting is the card terminals themselves. Meaning when you swipe your card, as the transaction is being conducted they take a copy of the information and re-transmit it to a host somewhere they control (or to a C&C machine on that network, which they then exfiltrate somehow).

Which is accurate:

> The apparent credit and debit card breach uncovered last week at Home Depot was aided in part by a new variant of the malicious software program that stole card account data from cash registers at Target last December, according to sources close to the investigation.

So malware was intercepting the raw CC information inside the card terminal on its way to the payment processor upstream. This allowed them to remotely infect every single Target store in the country, and steal information without setting foot there.


Ah, the distinction being that it wasn't stolen in a single bulk breach or via a device placed between the card and the scanner. Fair enough.

Yep, lots of these point-of-sale machines are (still!) running XP Embedded which is, AIUI, pretty heavily targeted. Apparently -- but not surprisingly -- they aren't updated all that much so they're pretty susceptible to being compromised.

Yes, for point-of-sale exploits, this would not work. But certainly there are companies that store CC information. Every time I go to Amazon, I don't re-input my cc info. They must be storing it somewhere?

As for buying it back, see my comment about LE not needing to make such purchases for this to work. And if the CC companies know about the fake numbers, the process of verifying that the number works, would itself, inform of the breach.

> Most large retailers only store the full credit card information for seconds while they process the transaction, when the transaction is complete it is discarded.

Couldn't work around that by having some device that attaches to a randomly selected POS terminal, and then executes a series of dummy transactions using thousands of "canary cards" that were assigned to the store?

They could perhaps set this up with the credit cards companies so any use of the canary cards are automatically flagged when they're used outside of the store that was issued them.

Could very well work, only you would need cooperation with one or more major payment processors/card issuers.

The way it would work is that shops are given hobeypot cards and asked to use them weekly for dummy payments. The moment one of these cards show up anywhere else than the shop it was sent to you can expect the shop has been hacked.

It is not failsafe and you will have to wait until a batch with this card is stolen before this tells you where.

>Plus even if you could get your "gotcha" credit cards into the criminal's hands, the chance of law enforcement buying one back is tiny, since the stolen batch is mixed with other cards, and only then a small sub-set is sold (since they verify if the card is still active, and meets other conditions).

You wouldn't need law enforcement to buy it. You just need the bank to watch for a transaction on any of these cards that's not from the merchant it was assigned to.

Bloomberg does this with data sales. They'll sell you particular financial information, like tick-by-tick for the Russell 2000 over the last 5 years. They embed unique hidden markers or fake trades with everyone they sell them to. They then go and acquire data from other "providers", and find the leak.

Can't people just make two independent purchases of the data and diff them to find the issues?

Would fake trades not be distorting the data you are purchasing from them?

From what I understand, they used to change volume slightly which they claim wouldn't negatively impact algorithms looking at historical data. Then, however, people were using volume weighted average pricing (VWAP), which would eliminate that change. This is just theory now, but I believe they just add double trades, instead of fake trades are certain prices.

You'd need to buy too many cards to find them by number, tho the combination of issuing bank, IBAN, card type, expiry date and [post|zip]code would likely provide enough bits to identify a source.

The shops would combat this by narrowing the information available about a card pre-purchase.

Here is what a record looks like on Rescator (a popular autoshop for credit cards) pre-purchase:


And to add, the credit card companies would know about the fakes so as soon as someone tried using it, we'd know about the breach and where it came from. All without having LE trying to buy the information from criminals.

Not directly related, but this reminds me about the story of paper town in map making. Check this TED's talk: http://www.ted.com/talks/john_green_the_nerd_s_guide_to_lear...

Pig notice only on checkout? That doesn't necessarily point to some issue with his payment method. The checkout server is likely different than the content server and may be subject to different protections.

Imho they are running the standard 'enemy IP' blocklists used by p2p filesharers for may years. Where one sees law enforcement using computers, one must first assume incompetence. In all likelihood he/she was using a machine at a known cop shop (Windows) without any sort of IP masking.

Some cops are highly trained IT specialists, but many investigators are actually just senior traffic enforcers who have been promoted into "technology crime" units. They have training and paper certifications, but lack any deep knowledge re technology. They make lots of mistakes. Think flowers by irene.

You missed an important point: if they wait until checkout then the payment process will complete and they'll get to walk away with law enforcement's money.

They have every incentive to wait until the very end to tip their hand.

These fraud shops make you deposit first then you go shopping much like how silk rd worked, so at anytime could have seized the coins.

For all we know this screen upon checkout is random, and whoever doesn't complain to get their money back is the cop. Most fraud outfits have major customers they deal with or cashout teams they offer specific BINs over chat, the online store is for bottom tier thieves on public sites.

I don't understand why they tip their hand at all. Why not just sell them randomly generated numbers?

If they really wanted to increase their take they could have that fed identifying system running on the content server as well. That way they can show inflated prices to feds, have the feds try to buy, then walk away with a bigger haul.

This may be an over assumption, looking at the timing of the block. Article states that because of the timing the site owners were able to seize his funded account, getting a few hundred dollars for free.

May have flagged him prior, but like the opportunity to take government money.

10 years ago, maybe, but nowadays federal-level LE runs extremely high opsec. Theres no way they would've run something like this from a government IP, let alone a non-virtualized machine.

But how many investigators are actual feds? There are literally hundreds, thousands even, of US police agencies. Most any one of those can setup a few computers and start looking for online crime. This guy could be from a local department of five officers, not an unusual situation in the US.

You would think that law enforcement would already be using a VPN or Tor for stuff like this.

It's entirely possible the LEO was masking his IP in some way; he didn't give any details. There are many other ways of flagging fraudulent (or anti-fraudulent...) accounts and transactions.

Could be blockchain analysis. Block accounts that receive bitcoin from known seized funds, like a Silk Road wallet. Basically just the opposite of tactics companies like Coinbase employ.

Thats what my bet is on. LE agencies have contractors that procure bulk quantities of prepaid phones, prepaid credit cards, etc.. for undercover investigations. I suppose its possible they traced the BTC from a funded wallet they knew was government related.

If they had prepaid credit cards, why wouldn't they just buy some Bitcoin from Coinbase with it? That wouldn't be traced back to them (unless another law enforcement agency was investigating... or if the fraudsters compromised Coinbase).

I feel like they'd at least be smart enough to use a tumbler if they were pulling it from known seized funds. They could also just give the investigator a $300 prepaid debit card card and have him buy some BTC on Coinbase.

The article speculates about some other possibilities.

> Then again, perhaps Rescator’s site simply noticed something amiss when my source funded his account with Bitcoin.

Maybe they walked the transactions back to some well-known funding source for law enforcement/government.

That's what I was wondering. The block chain isn't about anonymity, it's about validity, right?

I would think they'd block any IP that wasn't from a known VPN or Tor exit as a noob filter.

If criminals stopped selling to stupid people, they wouldn't have much of a market.

Smart crooks don't break the law. Stupid ones do.

Breaking the law isn't cost effective, and I am not talking about morals, I am saying that the additional overhead involved in covering your tracks and the potential liabilities (years of $0 or negative dollars income, legal fees) doesn't make sense.

Only rarely do you get a legitimately intelligent criminal, and that is often the result of personality issues/emotional problems.

This is nonsense. Without writing an essay here, you are apparently focusing on the criminals in jail.

There are plenty of very successful and intelligent criminals that have long, prosperous careers in theft.

Bernie Madoff ran a very successful, very illegal Ponzi scheme until he was nearly 80.

And that's not even to touch on the internet ability to let you violate the laws of counties you aren't physically in against citizens your gov doesn't care about (eg Russia to US), and therefore greatly reduce the risk for punishment.

I remember watching an episode of Lockup and realizing, hey, plenty of these imprisoned criminals are really clever -- they just use it for horrible purposes, like building makeshift knives to stab each other.

Noob dollars spend just as well.

Why are we still using Credit Cards as payment processing? Why can't we have something with a "Push" mechanism, instead of the "Pull" with the CCs.

I went to Rescator mentioned in the article and noticed they are using ICQ for contact..

Anyone know why?? Seems like an odd choice.

Tradition, all the Russian/CIS carders have always used jabber or ICQ mainly because they aren't worried about being caught and don't care about the security of their customers either. The bigger fraud outfits like Rescator are likely politically protected being a nephew to a cabinet minister or son of a Novosibirsk police captain. If you are a non connected Russian and try to run a fraud superstore the US feds can bribe lower level police agencies there to go make your life hell.

Your first sentence insinuates that Jabber is insecure. Could you elaborate on that? Just genuinely curious, since I used jabber/otr for a while to communicate with friends.

They seldom use OTR, I think infraud (in fraud we trust) still uses a group server in Azerbaijan to cleartext chat on none of these guys really care, some of the boards still use cloudflare.

Thanks for the clarification!

Icq is very popular on some countries, for example Russia.

There is a interesting phenomenon with IMs in that they tend to have popularity based on geographical locations and language borders, despite being on Internet.

Examples: in russia Icq is king. Brazil had Icq until MSN Messenger showed up, Icq tried and failed to get it back, gave up, then Microsoft became stupid and closed MSN Messenger, now the leader is Facebook messaging and whatsapp. Last I checked in us people used aim and yahoo clients.

Gmail messenger was a popular second place in many countries until Google killed it in favour of hangouts, now for text people use Facebook and for video they use Skype, that was, and kept their leading position.

In China the thing is Baidu client.

In Africa there is some messenger clients tied to local isps and phone companies.

I've never actually heard of successful anti-fraud busts.

Does law enforcement have any achievements or are they wasting time?

These don't even know to use vpn exit points and not use government ip addresses? Seriously?

They have lots of success, you just don't usually hear about it. Here is one of the few that actually got press: http://www.ecommercebytes.com/C/abblog/blog.pl?/pl/2011/11/1...

That's a really poor example of success.

The Romanian court handed down [a] suspended sentence

Because he already spent two years in prison.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact