Tag: compliance (Page 3 of 4)

PCI and the art of scoping

A lot of people we have met had told us this: “Since we are ISO27001, PCI should be a piece of cake, right?”

The context of this is because ISO27001 and PCI are often seen as distant cousins. They are both very relevant in our country and region (unlike other compliance like HIPAA), both deal with information security, and the overlaps between the Annex A controls of ISMS and PCI are evident. Therefore, the natural conclusion is ISO27001 is either a superset or a subset of PCI-DSS.

The problem with this assumption begins right at the start. In ISO, the scoping is largely determined by information that matters to your business. Before the iteration of 2013, scoping was generally done in a cowboy sort of manner. We met a company who had tons of sensitive information and told us they were ISO27001 compliant – and their scope was the security of their printing documents. Yes. How they literally secured the hard copies of the printouts. That was their scope statement.

Now 2013 version of ISO27001 had tried to stem these shenanigans by introducing interfaces and dependencies. Basically now the scope needs to cover information that are deemed important enough to protect from business perspective. This would cover the products and services relevant to the context of the organisation. Overall, scope determination of the ISMS can be a prolonged matter if you have a large organisation, and is often subjected to the business side of the organisation.

PCI scope?

It’s determined by the information that matters to the payment card brands, not your organisation. Credit Card Information. Primary Account Number. That’s it.

If you store, transmit and process PAN, PCI applies. If you don’t do any of these, and you do not influence any transactions in the payment card flow, then PCI doesn’t apply.

Often, people express utter shock that PCI doesn’t have any business continuity requirements. In terms of the holy trinity of information security – Confidentiality, Integrity and Availability (CIA), PCI primarily focuses on Confidentiality. Integrity is only focused in terms of its relation to confidentiality (Integrity of logs, integrity of system changes, system files etc), and there is no concern on Availability. Which makes sense. Between you closing down your business for one week versus you losing credit card information of your customers, the latter is viewed as more critical to the payment brands than the former. Although from a business perspective, a loss of business for a merchant is a loss of business for the entire data flow, upstream or downstream, so PCI not really caring about your RTO or RPO may be counter productive – but that’s an argument for another day.

At the end PCI scope boils down to you storing, transmitting or processing card holder data (CHD). Even if you don’t do any of these 3, you might still be in scope if you influence the security – an example would be those SAQ A e-commerce merchants that redirects requests to another PCI service provider. Even though they don’t deal with the CHD, they influence the transaction through their redirects, therefore, some parts of the requirements need to be met.

So – before we start our PCI journey, it is  very important to know what is the scope that is covered in your PCI environment. We may not want to take the whole environment as IN SCOPE – for cost, quality and timeline purposes. Our normal practice here is to reduce the scope as much as we can, a process we consultants term as “Scope Optimisation” simply because it sounds grand. I mean it sounds better than “Reduce your scope” which generally is interpreted to “reduce your price”.

In general, there are six things that we have to compile before we truly initiate the PCI journey.

a) Location and Address of the PCI scope. This is simple enough. Usually your data center is in scope. Depending on whether you store, transmit or process card data in your other offices, those come in scope as well. A question here would be – what about HQ, where our administrators access the PCI systems in DC, via a VPN connection? Ah. The secret sauce of putting things out of scope in a remote location where there is no storage, transmission or processing (lets just shorten these to STP from here on) of card data but there is access from an admin systems – multi-factor authentication. As long as this is in place, while the admin system is in scope, the location itself is then put out of scope. So you can connect from Starbucks, your home, or Timbuktu, and you would not have these locations dragged into your precious scope.

b) Applications that STP CHD. Store, transmit or process card holder data. Many queries have been like – oh, do we need to use PA-DSS applications? Well, if you do use PA-DSS certified applications, it would be very useful. However, even if you do not, you can still access that application as part of your scope under Requirement 6. In fact, some applications may not even be able to be PA-DSS for many reasons, such as it not being part of the authorisation or settlement flow but still storing card data. A custom CRM for example would be one that cannot be PA-DSS but still in scope for card data application. OTC (off the counter) products that store card data are still in scope, however, they need to be assessed properly to determine if there are any security issues that may influence the confidentiality of card information.

c) Network Diagram – an updated network diagram is a must. And a network diagram needs to be detailed enough to be able to differentiate the PCI and non-PCI zones. The important thing we need to take note on the network diagram is the proper demarcation of PCI zones, so we know what are:

  1. Card data environment in scope (CDE-IN-SCOPE) – ZONE A
    1. Any system that store or process or transmit CHD
    2. For example, application server, Database server
  2. Non-Card data environment in scope (NON-CDE-INSCOPE) – ZONE B
    1. Any system that require to communicate with CHD
    2. For example, patch server, anti-virus server
  3. Out of Scope – ZONE C
    1. System not related and has no communication with CHD – but might communicate with NON CDE IN SCOPE.
    2. For example, CCTV server in your office environment

d) Asset List for PCI – the asset list is critical because this relates directly to the effort and remediation costs of your PCI program. There is a huge difference in doing pentest for 200 systems versus 20 systems. So in this case, we don’t care about your assets considered not in scope, we want to know the assets in CDE and in NON-CDE in scope (Zone A and B).

e) Public IP addresses – this is needed because of ASV scans required. ASV scans are security scans done by the ASV (Approved Scan Vendors) of PCI. You can’t do it yourself, you need to get an ASV to do this for you.

f) Data Flow Diagram – This shows the card data flow in your organisation. Basically every channel where credit card enters into your environment, stored and process and exits. This details the lifecycle of CHD in your organisation whether it ends up being stored in a database for seven years, or passed out to another service provider. It’s essential to understand this – and if you have multiple channels where card is being entered (e.g e-commerce, POS, MOTO, Call Centers, KIOSKS etc) you need to document each of these from start to end.

So there you have it. PCI scoping at your fingertips. Drop us an email at pcidss@pkfmalaysia.com and we can have a free session with your organisation on what could be your possible scope, which likely may not be just your printouts coming out of your printer!

PCI-DSS and the Pervasive Certification Myth

The pervasive certification myth is so pervasive in PCI-DSS that we are going to give it its own Acronym: PCM. Because we are so tired of having to explain this over and over, we are going to canonize this corporate disease called PCM and forever immortalise it as the one of the most deluded, misleading and misinformed quackery to ever blanket the PCI-DSS industry, the same sort of quackery that insists urine therapy should still be used today for natural teeth whitening. Yes, it seems appropriate to compare these two in the same breath.

Please note that the below article is satirical (borne out by our immense frustration and oftentimes resignation that this will never be properly sorted out, ever).

PCI-DSS and PCM

The history of PCM has its roots in PCI-DSS itself being considered as a standard that can be certified against. Because of the name, Payment Card Industry Data Security Standard, immediately, fairly important people in the financial and banking industry who generally prefer to spend more time golfing and drinking fine wine than to actually read the standard and understand it better – these people concluded that like any other standard, there must be a certificate to ‘prove’ you are compliant. I mean, why not? ISO9001 has it. ISO27001 has it. Additionally, these were the same people who insist on having a certificate for literally everything that they do in their corporate life, from attending a half hour seminar talk on how to grow daffodils, to sleeping though a computer based training webinar; to providing a poor soul whose car has broken down some assistance in pushing it to the side of the road. Where’s my certificate to prove that I helped you, my good man?

Apparently, certificates are absolutely critical to our financial industry, without which our entire economy would surely descend into the Dark Ages of Financial Purgatory. In the same way, it has perpetuated into PCI-DSS that despite everything that the PCI council has said and has begged the entire industry to reject this PCM quackery, 99% of the time, we are still faced with the mind boggling, soul numbing, heart wrenching email or question stating: I want to see your PCI certificate.

Here’s the official statement from PCI council:

FAQ #1220

Are compliance certificates recognized for PCI DSS validation?

 

No. The only documentation recognized for PCI DSS validation are the official documents from the PCI SSC website. Any other form of certificate or documentation issued for the purposes of illustrating compliance to PCI DSS or any other PCI standard are not authorized or validated, and their use is not acceptable for evidencing compliance. The use of certificates or other non-authorized documentation to validate PCI DSS Requirement 12.8 and/or Requirement 12.9 is also not acceptable.

 

The PCI SSC website is the only source of official reporting templates and forms that are approved and accepted by all payment brands. These include Report on Compliance (ROC) templates, Attestations of Compliance (AOC), Self-Assessment Questionnaires (SAQ), and Attestations of Scan Compliance for ASV scans. Only these official documents and forms are acceptable for the purposes of compliance validation.

 

Because certificates and other non-authorized documentation are not officially recognized, entities that receive these documents to indicate their own compliance (for example, from a QSA or ASV) or another entity’s compliance (for example, from a service provider) should request that official PCI SSC documentation be provided. Any organization issuing, providing, or using certificates as an indication of compliance must also be able to provide the official documents.

OK.

We actually would prefer the answer to just be: “No. For heavens’ sakes stop asking the council such questions and waste our typing time on the keyboard!” But we are not the council. Else we will be out of work within a day.

You ask then: Wait a minute, if there is no certification, what are we supposed to submit then?

Well the answer to that is: it depends.

If you are a level 1 merchant or service provider, then you submit your Attestation of Compliance (AoC) signed off by yourself and the Qualified Security Assessor (QSA) or Internal Security Assessor (ISA). You may also submit your ASV scan reports, or if requested your full Report of Compliance (RoC). But rarely those last two are needed. What is needed is the AoC. That’s your PCI ticket. If you are doing a self signed Self Assessment Questionnaire (SAQ), you submit in your AoC, and if required the full SAQ documents with your response on the applicable questions.

I suppose we (the consultants, advisors, auditors) didn’t help in stopping this PCM disease ourselves, by oftentimes referring to Level 1 as a PCI ‘certification’. It’s more of figure of speech than anything, meaning that a third party is to validate your compliance as opposed to you doing a self validating process under the Self Assessment Questionnaire (SAQ) path. So we end up saying certificate, certificate, certificate and finally everyone asks, well, then, where is my certificate? As in the actual, real certificate, and not this figure of speech one?

It also doesn’t help that most (if not all) QSAs play along with this ridiculous fallacy. They will come up with their own ‘certificate’, made to look very official, very grand, very formal and very important looking. You know, those papers with flowery borders at the side and some huge bold statement saying, So and So are official certified, and signed off by a very important person. Some even put a seal in there for good measure. As if they graduated from Hogwarts.

The truth is, that certificate paper is just piece of paper. It’s not acceptable in any way for PCI-DSS and shouldn’t even be requested by acquirers or banks. Yes, no matter that the QSAs even make it look like it’s gold laced, and our customers print it out in all its Arial font glory, and spend a few bucks to frame it up nicely. That’s all well and good, and they are entitled to do anything with it, but PCM rears its ugly head when acquirers, banks, customers and clients insist on having it. INSIST.

If you have gone through the level 1 validation with a QSA and that QSA is able to provide one of these certificates, as a consequence of the PCM disease, then I suppose it’s fine, you can just play along as opposed to haranguing them about the PCM disease like what we are doing right now. Just provide them that dratted document. This does increase the myth further though, and it starts infecting the entire industry even more, because, now the acquirer/client/bank will go to another company and declare they need the certificate that the other merchant had provided. Until they reach a small company who is doing their own SAQ, whereby they say: “Yes, even if you are doing a level 4 merchant compliance, you should still be having a certificate! Come on now, chop chop!”

And that’s what we are facing.

Without exposing which industry right now we are assisting (those following our recent blog posts may venture a very educated guess), we are helping a lot of merchants undergo level 4 self signed SAQs. This is absolutely allowed by PCI-DSS. There is zero requirement to get QSAs involved. ZERO. In fact, PCI council printed this out in their Top 10 myths of PCI-DSS.

So now, our clients want to submit their compliance document (AoC) to the group requesting for it and here is the summary of response they received: 

a) The group requesting our client’s compliance says that the SAQ needs to be sent over to the QSA so that the correct SAQ type applicable to our client’s business will be determined. (WHAT?!)

b) Our client only now has to submit the PCI DSS compliance certificate.

I looked at it and we just shrugged. Resignation. PCM. This is what this disease has caused.

Firstly – the first phrase is proof that the writer has not bothered to even understand the background of SAQ and PCI.

No. The SAQ DOES NOT NEED TO BE SENT TO THE ASSESOR! It doesn’t even make sense. You are telling me to send the assessor the SAQ document I filled up so they can determine what SAQ I need to fill up? What sort of recursive devilry is this? Is this one of those tricky while-do loop we do as programmers that never ends because the condition to end is the condition to begin as well?

How many times have the PCI council stated that it is not the QSA’s role to define the validation requirements of the merchant, or the service provider or the bank. It’s NOT. The QSA’s role is to assess based on whatever validation requirements that has been determined. Yes, they can advice. Yes, they can with their amazing experience and god-like understanding of PCI-DSS, suggest which validation requirement to take. But at the end, the acceptable validation requirements are based on the bank, the acquirer, or worst case whichever company requesting the merchant to be compliant. If none, then the validation requirements must fall back to the guidelines provided by PCI SSC, which means, it falls back to the individual card brands. We are not going to go into that for now, but in general, VISA/Mastercard has an agreement on merchant levels, but Amex has some weird levels of their own. What we are stating is that, if a merchant is considered a level 4 merchant based on its volume, but the acquirer decides that they must still undergo Level 1 validation, then that’s the acquirer’s call. But it’s never the QSA/consultant to decide this. The QSAs job is to assess the company against the validation requirements and have an opinion if it is pass or fail. The type of validation is determined by the acquiring party. So, no the first sentence is already incorrect, never mind the recursive gibberish.

The second sentence is where PCM disease kicks in. Because nothing is known about the ‘AoC’, everyone immediately assumes that the certificate is an official document from PCI-DSS and it should be sent in. No. If the merchant is doing a level 4 self signed SAQ, that is 100% allowed by the council and by their acquirers, where in Thor’s Holy Hammer are they going to get a certificate to fulfill your insatiable lust for PCI certificates??!!

I am tempted to just have my 5 year old son draw a smiley face on an A4 paper, and stick one of his favourite Cars 3 character sticker on it, sign off and laminate it.

Go ahead – I dare you to google “PCI certificate” and click on ‘images’ and see the result of PCM in our world. Thousands upon thousands of PCI certificate documents, some even daring to put the PCI SSC logo on the certificate as if to say PCI SSC has endorsed the certificates, to provide these documents an air of official integrity. Even worse, we see QSAs giving ‘certificates’ for clients undergoing SAQs. Wait, if they have audited them, then OK, fine. But if is a self signed SAQ, how can certificates be even provided?

We are not saying what they are doing is wrong. No, it’s not wrong to provide a PCI certificate. But PCI SSC has clearly stated that if you guys want to do this, you must state clearly that these are not official PCI documents and these are supplementary, not mandatory and only provided by the QSA/ASV and not endorsed by PCI-SSC and cannot replace the official documents like AoC or RoC. Basically, PCI is just saying, “Put it up in your office or your lobby so you can brag, but please don’t show it to us and say you are PCI compliant, we rather be looking at a piece of art written by a 5 year old kid with some Cars 3 character sticker on it.”

Finally, to end this article (or rant, it may seem to some). The reason why this is written is to drive home the fact that PCI-DSS has no actual paper certificates. None. Whatsoever. The actual document you should be requesting for is the Attestation of Compliance (AoC). Please do not ever request for the Certificate of Compliance ever again because this means, you are guilty of spreading the epidemic of PCM across the world.

Note: This article is meant to be satirical, for us to blow off steam and not intended to offend any party or to dispense actual advice. There is actually no such thing as an official PCI Pervasive Certification Myth. At least, it has not yet been officially defined, as far as we know.

 

 

 

IATA PCI-DSS: New FAQs!

So, it has been a while since we’ve updated on the ongoing PCI-DSS program from IATA. Just a brief recap then: Airlines have demanded that IATA support their own internal compliance project by making the BSP (Billing Settlement Plan) card sales channel PCI DSS compliant. This is why IATA Accredited Travel Agents now need to become PCI DSS compliant by 1st March 2018. Yes, that’s roughly 6 weeks ahead of this writing. And no, it doesn’t seem like there might be any extension towards this compliance from IATA. However, there are some pretty big news headed your way on this compliance, as we are in touch with IATA over the last couple of months and also assisting many travel agencies to get PCI-DSS sorted out in their payment channels.

However, for this article, we will focus on the brand new FAQs that just came out a few days ago (18 Jan 2018)! You can find the updated FAQs here at http://www.iata.org/services/finance/Documents/pci-dss-faqs.pdf, and we are going to look through a few changes.

FAQ #3

What if I do not have an acquirer?

Old FAQ: We suggest that you contact the credit card branch that you are working with.

New FAQ: In that case, you are solely accountable for the PCI DSS compliance of the BSP card transactions you are making on account of the airline whose ticket you are selling. We suggest you contact your GDS provider who can provide some guidance, and then review through which of your systems card details transit or are stored. Starting from this you will know which of your systems
must undergo a PCI DSS evaluation.

Our opinion: The first FAQ was of course, not exactly extremely helpful, since most credit card branch does not give two hoots about travel agencies banging down their doors in search of their response. The new FAQ is basically saying, well – you just need to figure out yourself then, but you can ask the GDS guys if you wish. We have. The GDS guys are very important in this factor, because they first need to be PCI compliant. Sabre, Amadeus and I think Galileo Travelport is. Secondly, they can give some guidance on how agencies can approach PCI based on the client software that is installed on the agency side.

What do we mean by this? Because for agencies not storing credit card, they can possibly be eligible for shorter SAQ (Self Assessment Questionnaires) for PCI. An SAQ D has 340+ questions. An SAQ A has only 20+. If an agency uses the GDS for credit card passthrough transactions (i.e the credit card form of payment), and not store credit card information in the back office or any electronic form (email, skype, excel etc), they might qualify for shorter SAQs. The question is which?

Some advisors claim the SAQ C is correct due to the fact that the GDS is a payment system. The reasoning is that this is no different from integrated POS systems like Micros. In Malaysia, we have hundreds of different vendors in POS solutions for retailers, F&B franchisees etc. But is the GDS really like an integrated POS solution? SAQ C has around 160 questions. The amount of time you will spend on this is probably the same amount of time taken to watch two seasons of the Game of Thrones. Or three, depending on whether you binge watch or not.

Some advisors veer to the other extreme, claiming that the GDS client is simply a browser system that is redirecting the entire card data processing work to the GDS provider, so they are eligible for A. 22 questions. Maybe an episode of Seinfeld. But A is generally for a web browser based site with absolutely zero handling of credit card on their end, not just systematic, but also manual. The only way this works for travel agency is that they outsource an entire call center to handle their MOTO business and do not accept walk-in customers. I don’t think that’s happening. Most feedback I get from livid agencies about PCI-DSS is that they are struggling too much on thin margins. So, no, SAQ A is entirely too liberal.

SAQ C-VT has a seemingly better balance to it, as discussed in our previous articles Part 1 and Part 2.

We even sent out queries to two GDS (their names pending once I get their agreement to publish) and their responses were these

Amadeus: (When Queried if SAQ C-VT is correct to be filled, and if the Amadeus Selling Platform can be eligible for VT): Basically, if the payment is done via Amadeus and entered manually from a personal computer directly into the GDS – you have a right form for Amadeus agents and tick it off with confidence. 

I believe your original question was ‘If Amadeus is considered virtual payment terminal?’

Our answer is Yes.

Sabre: (When asked if their client acts as a VT, defined by PCI as having “Internet-based access to an acquirer, processor or third-party service provider website to authorize payment card transactions.”) Yes, Sabre Red Workspace client requires an internet connection to authenticate and then it requires connections (dedicated or ISP with VPN) to connect to Sabre and no, it does not do batch processing. You may consider SRW is a virtual terminal and guiding your travel agency clients to achieve their goal.

Travelport (Galileo):  (When asked if their client acts as a VT, defined by PCI as having “Internet-based access to an acquirer, processor or third-party service provider website to authorize payment card transactions.”)

Yes. Galileo client does not store credit card information on the client software and client software requires internet connectivity, and cannot do batch transactions.

Based on these ‘guidance’ from GDS which IATA seem to defer to, SAQ C-VT is a likely possibility, as long as all the other eligibility are met. The GDS all claims they are virtual terminals, but that itself (while an important eligibility) isn’t the ONLY eligibility for SAQ C-VT, so you need to ensure the others are met before claiming SAQ C-VT is correct or your business.

Whew. That was a long one. Now back to our FAQs.

FAQ #9 : As a travel professional issuing and selling airline tickets, am I considered a merchant?

This is removed and rightly so. Though the previous response was right: “All the airline transactions processed through a GDS (Global Distribution System) and IATA BSP, the airline itself is considered as the merchant, not the travel agent.”

It only serves to confuse an already confused population further. It’s better they don’t explain this, because some agencies interpret this as IATA saying they are not ‘merchants’ so they need to be ‘service providers’. WHAT! So, yeah, we can explain in another article but this is better left out.

FAQ #22: We already have a PCI DSS Compliant certificate issued by a third party.
Is this enough to cover our BSP or do we need to complete more forms?

Not an addition or whatever, but I still wish that they would change this because the answer doesn’t match the question. The answer is lifted directly out of the PCI-DSS Top 10 Myths addressing the need for a QSA to be involved in the process. The answer is , it is recommended, but NO, for Level 3 and 4 merchants, there is no requirement to get a QSA involved.

Finally, a bonus opinion here.

Many agencies are still faltering in their PCI-DSS compliance. Some equate that just because they are level 3 and 4, they do not need to do ASV scans or penetration testing. Likewise, there are those who *might* theoretically (we don’t know any) qualify for level 1 or level 2 based on their volume, automatically assume they need to do ASV scans and do pentest for everything in scope.

NO.

Your merchant level DOES NOT dictate whether you need to conduct PCI scans or not. We need this to be clear. Because the table published in the FAQ from IATA for FAQ#13 isn’t clear (not their fault, this was lifted from the Mastercard site) – the column “Validated By” states ‘merchant’ and below “Approved Scanning vendor” for level 2 and below. This immediately presupposes that an ASV must be involved. This is incorrect.

Your level (determined by your card transaction volume) determines your VALIDATION TYPE. Validation type there are 3: QSA Certified/Validated; Validated SAQ by QSA/ISA and SELF SIGNED SAQ by MERCHANT OFFICER. That’s it. Your level doesn’t determine how you go through PCI, it determines how it is validated. And it’s not set in stone. Your acquirer can bypass these guidelines and decide that even if you only do ONE transaction a year, you still must go through level 1 compliance (audited by QSA). This is actually quite common!

So what actually determines what on earth you actually do in PCI-DSS?

Well, it’s your business. Or, for Level 2 merchants and below, your type of SAQ. You see, it’s your business that determines your SAQ type, it’s your SAQ that determines what you need to do, and based on what you have done, it will be validated in either of the 3 ways we’ve described above. That’s the harmony of PCI. That’s the zen. The yin and yang. The balance in the Force.

So, for instance, if you are doing SAQ A, SAQ B or SAQ C-VT, please point out to us the fact that you are REQUIRED to do ASV scans on all your internet address (some are told, even their dynamically allocated broadband IP must be scanned by ASV).

None. Magically, SAQ A, SAQ B and SAQ C-VT DOES NOT HAVE ANY requirement for ASV or penetration testing. For us who can provide these services, of course it kind of sucks since now those going through these SAQs don’t need our services anymore. But we rather tell them straight the correct way and sacrifice that part of our business than to let them know wrongly and give consultants a bad name. So what SAQ you are doing will determine whether you need to get something scanned or not.

Now, of course, do not be tempted to fit your business into the easiest SAQ for the sake of it (see the example of travel agencies with GDS doing SAQ A) – there are huge eligibility requirements for these 3 SAQs and not many agencies can meet it. If you practice accepting cards through email, or photos on Whatsapp for your credit card; or store in back office for later processing, or have Enhanced Data Services from Visa/Mastercard or a thousand other ways you can be receiving credit card, you likely need to fit back into the dreaded SAQ D. But what we are saying is that if you ARE eligible for A, B or C-VT, then those will determine whether you need to do any testing or not.

It is our opinion that testing and scans should be done regardless for security sake, not so much for compliance but the choice is yours. You need to make that decision for your own business. Because that’s what heroes do.

If you have further queries on PCI-DSS or just how we are currently helping our clients get through PCI, drop us an email at avantedge@pkfmalaysia.com. We will respond ASAP!

PCI-DSS IATA: Dissecting the New FAQs

A few significant things occurred this week for the IATA PCI-DSS Program, summarised below:

a) We finally have a very clear way forward thanks to some clarifications direct from IATA, and in some parts due to our dogged persistence to get some answers

b) The new FAQs were published end of June and an updated version was done yesterday (11 July) and is now online at http://www.iata.org/services/finance/Documents/pci-dss-faqs.pdf

Firstly, the significant news.

IATA confirmed that Level 3 and Level 4 Merchants do not need a QSA to signoff their AoC/SAQ – which, to many agents, means they can do SAQ on their own, or using their own IT resources, or external consultants (not necessarily QSA, but if you prefer a QSA, by all means, go for it)

IATA also confirmed that they are considering exemptions for agencies that do not have any credit card transactions in their business channels.

These two clarifications address some long running questions agencies had for PCI-DSS. Do they need external consultants, do they need a QSA, do they need any compliance even if they don’t have credit card, etc etc.

Regarding point b) above, there was a quick iteration on the FAQs to clarify a few items. So here are some of the changes between the newest FAQ on 11 July and the one on the 29 June, and we can go through it.

The first 4 FAQ questions remain more or less the same although we do have a nitpick on 1, which is

FAQ#1 Who do I approach for PCI DSS compliance?
We suggest that you contact your acquirer.

Technically, this is correct, however, it’s not exactly complete. Because their (travel agents) acquirer wouldn’t have visibility over the agencies’ channel of credit card via GDS and BSP (or soon to be NGI – the new gen ISS). Acquirers have no idea of this because when the agents uses GDS credit card facility, they are doing in BEHALF of airlines! So even if they were to correspond with the payment brands directly as per FAQ#3, the brands wouldn’t know, nor care about the agency. Because in the GDS-BSP channel, the agency is not the merchant – it is the airline. (lightbulb).

Therefore, it must be the airlines who must be PCI compliant in that channel – however, because they make use of agents, the agents end up having to be compliant as well. But the airlines don’t deal directly with agents for this channel – they have an aggregator in between the agencies and airlines. And yes, this aggregator, this glue that holds everything together is the ecosystem of GDS-BSP/NGI. So if the agency connects to BSP, IATA is the ‘service provider’ offering this service – therefore, it is IATA that needs to clarify the requirements. Which they are doing – so technically FAQ 1 should read

FAQ#1 Who do I approach for PCI DSS compliance?
Yo, it’s us, man! That’s why you’re reading this on an IATA page!

In our clarification request, we didn’t point this out to IATA because our email at that point was already too long. It’s like we were writing the Titanic of emails, and we had to cut some scenes to fit into a readable email size.

Next, FAQ #6 is also important. Only for our own selfish self satisfaction.

FAQ #6 Are compliance certificates recognized for PCI DSS validation?
The answer to this question is no. Any sort of documentation which is not under the authority and validation of PCI DSS, will not be accepted for indicating the company’s compliance with PCI DSS.

And this is what we have been telling clients for YEARS. There is literally no such thing as a certification of compliance as far as PCI is concerned. Yet, everyone wants to see your ‘certificate’ and even go as far as to reject the AoC and RoC and SAQ documents. There is NO SUCH THING as a PCI-DSS compliance certificate. If someone prints a certificate out for you, it cannot bear any logo from the PCI-DSS council because it is not part of PCI. It’s a nice piece of paper to put up in your lobby but that’s it. When we work with our principal QSA, they also have this “certificate”, but we always make it clear that this is only issued as an aesthetic by the QSA and not considered acceptable to the PCI-DSS program formally. You MUST have the AoC and RoC/SAQ combination of documents at least – and also whatever ASV scans etc you might have. So, we would suggest not to go about calling yourself PCI-certified agency – just say you are compliant to PCI is enough. It sounds less sexy but those are more accurate terms to use.

FAQ#7 was corrected to refer to Question 14, instead of Question 13 as previous FAQ stated. Innocent error, of course, no harm done. It doesn’t mean that the writer of the FAQ can’t count.

FAQ#8 was also corrected whereby the previous FAQ stated (emphasis ours)

“The latter has to be completed as a declaration of the results of the service provider’s assessment with the Payment Card Industry Data Security Standard Requirements and Security Assessment Procedures (PCI DSS).”

and now correctly states

The latter has to be completed as a declaration of the results of the merchant (or travel agencies)’s assessment with the Payment Card Industry Data Security Standard Requirements and Security Assessment Procedures (PCI DSS).

Because technically, travel agents are merchants not service provider, so it might be just a copy and paste error.

FAQ#11 Can a QSA that is not listed in a specific country but listed in another country conduct a certification process in the non-listed country?

Originally it stated

“Yes. By definition, Qualified Security Assessor (QSA) companies are independent security organizations that have been qualified by the PCI Security Standards Council to validate an entity’s adherence to PCI DSS.”

This might not be exactly accurate as there are certain regional restrictions based on fees. So, the change became

“Overall speaking, yes. Nevertheless it should be noted that under the QSA program guide, section 6.3.1, there are qualified regions in which QSA can or cannot perform in as noted “QSA Companies are authorized to perform PCI DSS Assessments and QSA-related duties only in the geographic region(s) or country(s) for which they have paid the regional or country fees, and as indicated on the QSA List.”

Once again, that’s more accurate. And there are more words. And it has a quote from an official document from PCI, so it sounds very important.

FAQ#14 has the big change in the new FAQ version compared to the one in June, whereby under level 2 merchant column we have this “note” to clarify that level 2 merchants under Mastercard requires either an onsite ISA or onsite QSA to validate their SAQ. This is what we call “Validated SAQ”, and this is what Mastercard was telling us earlier, that this must be done by the QSA onsite (if they do not have an ISA – which is “internal security assessor”, which is as rare as an albino beluga whale).

FAQ#18 is the one applying to agencies without any credit card transactions. Now you do need to be careful. IATA does state that PCI applies to agencies processing credit card with the IATA GDS-BSP channel or any other channel (including your acquiring bank direct channel). This means if you have an EDC or POS device, or do internet transactions, you STILL need to undergo PCI-DSS. Who you send your AoC/SAQ to is another story, because IATA wouldn’t know much about your POS/EDC channel since you are not their merchant. Technically you send it to whoever you have a merchant account with – your acquirer. But again, your acquirer isn’t even asking for it! So. We suggest that you still do it, and keep it in case someone asks for it. We hear some cynical snorts in the background but we are going to ignore it. Be nice.

FAQ#23 – they decided to completely change this one. The previous answer seemed slightly confusing and in contradiction to FAQ#6 and FAQ#14. Previous FAQ in end June stated:

It should be noted that the third party should be authorized by the PCI DSS Council as a Qualified Security Assessor (QSA) to accept the PCI DSS compliance certificate. The scope shall cover the BSP card sale transactions.

a) Again, as FAQ#6 already confirmed – there is no such thing as a compliance certificate, so technically all compliance certificates issued by whoever, whether QSA, ISA, consultant or the Queen of England should not be accepted as formal documentation of PCI. One more time we hear this certificate of compliance being bandied around like its some sort of Ark of the Covenant, we are going to collectively walk out of our office and lie down on the main road in silent protest.

b) It’s sounds slightly confusing because it seems that this statement is saying a QSA is needed to be involved for all merchant level compliance as well which is contradicting FAQ#14.

To give them credit, their explanation to this was:

“We have had instances in which the agent was providing us with some sort of certificate issued by a third party, under the assumption that the certificate was issued by a QSA therefore we wanted to make clear that in the case an agent were to go this way they should be checking out the authorized QSA list available in PCI DSS council site.”

Yes, completely agreed. But not the certificate part. D@mn it, that’s it! We are headed out tomorrow and lying down on the street in protest!! Watch the news!

Anyways, now the current FAQ#23 reads a different:

Because most large merchants have complex IT environments, many hire a QSA to glean their specialized value for on-site security assessments required by PCI DSS. The QSA also makes it easier to develop and get approval for a compensating control. However, for Level 3 and Level 4, PCI DSS provides the option of doing an internal assessment with an officer sign-off if your acquirer and/or merchant bank agrees. Mid-sized and smaller merchants may use the Self-Assessment Questionnaire found on the PCI SSC Web site to assess themselves.

If this sounds familiar, it’s because it is. It’s lifted from Myth 6 in the famous PCI document at https://www.pcisecuritystandards.org/pdfs/pciscc_ten_common_myths.pdf

It’s certainly reads better, although, it doesn’t really answer the FAQ#23’s questions, but hey who cares? It makes sense!

So anyway, from this we learnt a few things

a) IATA is really keen to be on the ball on this PCI-DSS compliance and has put in effort in getting the right information out to the agencies – kudos to them and really impressed with their management’s response to queries.

b) The industry as a whole is still grappling a lot on PCI-DSS and needs to move forward with the right information and decision.

c) As QSAs, consultants, auditors, advisors or IT experts, we all need to work together to get our clients up to speed with the right information so they can make decisions and we can assist them.

PCI-DSS is never easy. Even those doing SAQ A-EP are having headaches, what more agents going through SAQ D-MER and all its 340++ questions. IATA seems to understand that and has PCI on their agenda.  We are willing to work with anyone on this – we have our clients who are travel agencies, but we also want to help other agencies get up to speed with PCI, what is required, and how to get compliant from the different validation requirements per PCI’s standpoint.

So to summarise this long winded post, from the horses’ mouth themselves:

a) There is no need for QSA involvement in Level 3 and level 4 merchant self assessment questionnaire (SAQ). Merchant officer signoff on section 3b is enough. However (and this is our opinion) if you can get assistance from QSA, ISA, consultants, IT experts, auditors,the Queen of England or even your own internal IT person familiar with PCI, go for it. You’ll need all the help you can get.

b) For those without credit card transactions in ALL channels (not just IATA), consider the exemption in FAQ#18. But please contact IATA on this as you should truly understand what might be the consequences in the future.

OK, that’s it for now. Drop us a note at pcidss@pkfmalaysia.com. We are preparing a complimentary talk on PCI-DSS specific to this travel agency industry soon, so stay tuned!

IATA PCI-DSS: Who is doing what?

pci-compliance

We have been receiving a ton of emails and inquiries lately ever since we started marketing our services for PCI-DSS to travel agencies. It has to be noted that some of these travel agencies were our clients to begin with. PKF has a very large set of customers because we do the entire end to end corporate services. We are not just one technology advisory firm. We have tax advisory, business advisory, internal audit, external audit, outsourced, accounting, corporate finance, forensics accounting etc. So over the course of 20+ years we have amassed a ton of clients and many of them are travel agencies, whether in technology group or in others. This is where our main queries stem from. Existing travel agencies are querying us and in turn they are letting others know about us, so much so that we are now compiling an FAQ to address all questions being thrown at us on PCI-DSS.

One common question we get asked is: WHO is initiating this PCI-DSS?? We even get accused of being the ones initiating this PCI-DSS on them and planting a deadline of March 2018 for them.

So let’s get the story here straight. For this, it is necessary to go from the beginning to the brief history of PCI.

a)  PCI-DSS began its life in 2004 but only in 2006, PCI Council was formed to govern this standard. The council is now made out of card brands Mastercard, Visa, Amex, JCB and Discover/Diners. The purpose was to ensure there was a standard way that merchants/service providers can secure their credit card interacting systems to, instead of to each individual card brand’s compliance. It’s a good thing. Basically the whole idea is to ensure the whole ecosystem where credit/debit card is used/processed/stored/transmitted is secured.

b) IATA’s story probably began back in 2015 when, according to GDS Amadeus, VISA Europe issued a deadline to acquiring banks using its network that all airline merchants should be PCI-DSS compliant by 31 December 2017. So the airlines got into a huff and took a look at their processes, which is like any other merchant – they have their acquiring bank to do the authorisation, clearing and settlement. So far ok.

c) However, the airlines had one problem: Indirect distribution channel. This is where airline tickets are distributed via travel agents, either through walkin, MOTO or internet. Travel agents use a GLOBAL DISTRIBUTION SYSTEM (GDS) that link to airlines to check for ticket and also to financial institutions for authorisation. And these finally link to IATA. Why? IATA has the Bank Settlement Plan (BSP) to – yup you got it – facilitate the clearing and settlement. BSP allows many travel agents to connect to many airlines, allowing a one stop shop to ensure everyone gets what they want, instead of travel agents separately dealing with airlines and vice versa. It’s orderly and it helps the industry.

d) However, the BSP, due to its connectivity to the Airlines now needs to ensure its downstream connecting parties are also PCI-DSS. Cue, travel agents and this is where IATA tells the travel agents, look, get your act together because the airlines need to be certified, so we need to be compliant, so you need to be compliant.

So in conclusion, it is IATA initiating it to the agencies – because there is an upstream push for them to be compliant. It’s common as well – many times payment gateways are asked to be compliant by their bank – we hardly see any entity embarking on PCI-DSS just because they feel that it’s the best thing to do for them. But the overall initiator of PCI still remains the card brands – whether it is VISA, Mastercard or Amex etc.

Now the question here is this – because IATA is considered a processor (with their BSP), they are enforcing a deadline of March 2018. At the same time, they also need to provide a way for agencies to submit the compliance document.

It’s a bit confusing here, because Agencies are also merchants in their own sense. They also have their own channels to collect payments, and some payments are made directly to their merchant account, and they settle with IATA through cheque/cash/bank in etc, not via card. Everytime a card is entered in the BSP, the agency is acting in behalf of the airlines using the airlines merchant ID. Everytime a card is used in the merchant’s own environment such as POS, EDC or Internet, the agent is the merchant, and they do authorisation, settlement etc through their own bank. IATA/BSP is not involved in the credit card flow in this case.

However, because IATA is requesting PCI to be adopted by agencies, agencies also need to look into their other channels that do not involve IATA! So imagine, an agency does their SAQ C/C-VT and sends it over to IATA, but to cover their EDC or terminal business, they do an SAQ B – who on earth do they send this over to? Well they send it over to their own acquiring bank. Their bank asks: Hey, what the heck is this? Well, it’s our PCI Compliant SAQ/AoC, Mr Bank. And Mr Bank is happy but somewhat confused and asks: Why are you doing it anyway? I didn’t ask you to do it yet because you only do 1 – 2 transactions with us. (Please note, even if it’s 1 – 2 transactions, you are still considered a Level 4 merchant, but most banks are ensuring their large volume merchants are compliant first). So therefore, agencies have two upstream processors to send their PCI documents to – IATA (for IATA channel) and their own bank, for others.

In the next post, we will explore on the validation requirements and why its so important to know what validation requirements apply to you and how. Do drop us a note at pcidss@pkfmalaysia.com. We are having a bunch of queries, but we will answer you ASAP.

« Older posts Newer posts »

© 2024 PKF AvantEdge

Up ↑