October 1st, 2015, is the deadline for merchants to upgrade “Point of Sale” and “Point of Swipe” terminals to recommended EMV compliant systems. To quote Wikipedia, “EMV (Europay MasterCard Visa), is a technical standard for smart payment cards and for payment terminals and automated teller machines which can accept them.” These new terminals can validate an EMV specific chip in a customer’s credit card on swipe, or validate a secure element in a mobile device when it is scanned by a terminal. The press is calling this transition “The EMV Liability Shift” because merchants who do not adopt the new standard for payment terminals are being told that they – not banks – will be responsible for fraudulent transactions. There are many possible reasons for this push.
But why should you care? I know some of you don’t care – or at least don’t think you should. Maybe your job does not involve payments, or perhaps your company doesn’t have payment terminals, or you could be a merchant who only processes “card not present” transactions. But the reality is that mobile payments and their supporting infrastructure will be a key security battleground in the coming years.
Talking about the EMV shift and payment security is difficult; there is a lot of confusion about what this shift means, what security is really being delivered, and the real benefits for merchants. Some of the confusion stems from the press focusing on value statement marketing by card brands, rather than digging into what these specifications and rollouts really involve. Stated another way, the marketed consumer value seldom matches the business intent driving the effort. So we are kicking off this new research series to cover the EMV shift, its impact on security and operations for merchants, and what they need to do beyond the specifications for security and business continuity – as part of the shift and beyond.
Every research paper we write at Securosis has the core goal of helping security practitioners get their jobs done. It’s what we do. And that’s usually a clear task when we are talking about how to deploy DLP, what DAM can and cannot do, or how to get the most out of your SIEM platform. With this series, it’s more difficult. First, payment terminals are not security appliances, but transaction processing devices which depend on security to work properly. The irony is that – from the outside – technologies that appear security-focused are only partially related to security. They are marketed as security solutions, but really intended to solve business problems or maintain competitive advantages. Second, the ecosystem is highly complex, with many different companies providing services along the chain, each having access to payment information. Third, we will discuss some security issues you probably haven’t considered – perhaps in the news or on the horizon, but likely not yet fully in your sphere of influence. Finally, many of the most interesting facets of this research, including details we needed to collect so we could write this series, are totally off the record. We will do our best to provide insights into issues merchants and payment service providers are dealing with behind the scenes (without specifically describing the scenarios that raised the issues) to help you make decisions on payment deployment options.
To amass sufficient background for this series we have spoken with merchants (both large and mid-sized), merchant banks, issuing banks, payment terminal manufacturers, payment gateway providers, card manufacturers, payment security specialists, and payment security providers. Each stakeholder has a very different view of the payment world and how they want it to work. We remain focused on helping end users get their (security) jobs done, but some of this research is background to help you understand how the pieces all fit together – and just as importantly, the business issues driving these changes.
- The Stated Goals: We will set the stage by explaining what EMV is, and what they are demanding of merchants. We will discuss how EMV and “smart card” technologies have changed the threat landscape in Europe and other parts of the world, and the card brands’ vision for the US. This is the least interesting part of the story, but it is necessary to understand the differences between what is being requested and what is being required – between security benefits and other things marketed as security benefits.
- The Landscape: We will sketch out the complicated payment landscape and where the major players fit. We do not expect readers to know the difference between an issuing bank and a merchant bank, so we will briefly explain the major players (merchants, gateways, issuers, acquirers, processors, and affiliates); showing where data, tokens, and other encrypted bits move. We will introduce each party along with their role. Where appropriate we will share public viewpoints on how each player would like access to consumer and payment data for various business functions.
- The Great EMV Migration: We will discuss the EMV-mandated requirements in some detail, the security problems they are intended to address, and how merchants should comply. We will examine some of the issues surrounding adoption, along with how deployment choices affect security and liability. We will also assess concerns over Chip & PIN vs. Chip & Signature, and why merchants and consumers should care.
- The P2P Encryption Conundrum: We will consider P2P encryption and the theory behind it. We will consider the difference between theory and practice, specifically between acquirer-based encryption solutions and P2P encryption, and the different issues when the endpoint is the gateway vs. the processor vs. the acquirer. We will explain why P2P is not part of the EMV mandate, and show how the models create weak links in the chain, possibly creating liability for merchants, and how this creates opportunities for fraud and grey areas of responsibility.
- The Tokens: Tokenization is a reasonably new subject in security circles, but it has demonstrated value for credit card (PAN) data security. With recent mobile payment solutions, we do not see new types of tokens to obfuscate account numbers or other pieces of financial data. We will briefly compare tokenization in merchant vs. banking systems, show how PAN data enters the system, and how it is replaced with tokens. There are three main deployment models: on-premise, Tokenization aaS, and third-party interception. We will explain how this improves security and helps reduces compliance burden and liability. We will review the impact on analysis and anti-fraud measures. Tokenization also impacts merchants, repayment, dispute resolution, and has produced services to address these requirements. We will review how Apple Pay brought tokenization to the attention of consumers, and largely blind-sided the industry. We will discuss the consumer side of payment systems; as well as how the model works, how tokens are created, where PAN data is stored, and how it fits in with merchant systems. This alternative approach brings new wrinkles to payment tokenization. The new mobile platforms and applications bring new risks to merchants, which must be considered when rolling out mobile payment solutions.
- Mobile Payments: We need to briefly discuss the principal security components for mobile payments, and perhaps just as importantly the operational adjustments needed to support mobile payments. We will review the Apple Pay and Starbucks mobile payment hacks, and the need for a fresh look at non-technical issues.
- Who Is to Blame? We will briefly address the liability shift and what happens when everything goes wrong, contrasting EMV against non-EMV deployments. Card brands offer a very succinct message to merchants: adopt EMV or accept liability. This is intended as a simple binary choice, but liability is not always that clear. We will explain the merchant liability waiver, how deployment choices help determine who is really responsible, and how liability is still an open question for some.
This research project was originally intended to be a short, focused look at EMV and the need for point-to-point encryption, but the investigation has produced some of our most interesting research over the past several years, so we will cover various related areas. Stay tuned for our next post, which will cover EMV’s goals.