5 Top Themes from MRC Vegas 2016
By Sripad Sriram /
21 Mar 2016
MRC Vegas is always a whirlwind – between keynotes, panel discussions, conversations at the booth, and lunchtime chats, it can be hard to keep up. Didn’t get a chance to attend the conference? Here are a few of the topics that kept coming up:

Internet of Things
There are two main trends making “the Internet of Things” a hot topic in the world of fraud. First, home appliances and digital devices like fire alarms, thermostats, watches or glasses are all becoming Internet-enabled and connected. Second, large machines like automobiles or aircrafts are essentially becoming digital devices themselves as technologies like autonomous vehicles diffuse into the marketplace and as companies like Apple start building cars themselves.
All of these connected devices just give bad guys more and more opportunities to defraud people. Often quoted at MRC was the case of fraudsters being able to remotely hack into the onboard computer of certain Jeeps, allowing hackers to stop a car while it was driving.
The question is how will fraudsters exploit the increased connectedly in our lives, and how can ordinary consumers protect themselves? What role do device manufacturers play in this, and how can vendors in the fraud industry help?
EMV
Although US merchants are now liable for card-present credit card fraud if their point-of-sale system isn’t EMV-enabled, merchants are taking their time converting from “swipe-and-sign” to EMV.
This affects fraud in a few ways. First, fraudsters still have a chance to acquire large quantities of U.S. card numbers by attacking “swipe-and-sign” POS systems while merchants still use them.
Second, omnichannel merchants (those that sell both online and in brick-and-mortar stores) still have a chance to put measures into place to protect against the inevitable shift from card-present fraud (i.e., using a stolen credit card in a store) to card-not-present fraud. Speakers often mentioned data from Canada and UK to support the hypothesis that CNP fraud will increase when EMV is more embraced by merchants.
Innovation helps bad guys as much as good guys
Technological advances bring us wonders like genome sequencing, self driving cars and the like. But the problem is fraudsters and criminals have the same access to all this new technology as the good guys, and they have the technical chops to use emergent technology for nefarious purposes.
Examples provided by Mark Goodman, author of Future Crimes, included Ukrainian start-ups that focused just on phishing and hacking attacks, drug dealers using Square to accept payments, and the more recent attack on Jeeps. Read more in our recap of Goodman’s fascinating talk.
Machine learning, deep learning, & AI
Machine learning has gone from an obscure technology used by research labs to something that fraud and risk managers advocate for and incorporate into their daily workflows. The main reason industry experts like Mark Wallick from Google recommend machine learning is to help scale fraud prevention. Legacy, rules-based systems just can’t keep up. At about 50–100 rules, adjusting rules to keep up with fraudsters is unmanageable.
Diving into the nitty gritty, there was some technical discussion of what specific methods were best for catching fraud, for example supervised vs. unsupervised machine learning. At Sift Science, we use supervised machine learning because we believe it’s a far better way to predict and catch fraud from nuanced data. As machine learning becomes more mainstream, anti-fraud vendors have begun exploring what the future holds. At MRC there were vendors discussing the merits of unsupervised learning techniques (like clustering) or deep learning to predict different types of fraud never seen before.
Increase friction for bad guys, decrease friction for good guys
Fraud teams are often considered the red-headed stepchild by merchants. These are the teams that may point out potential risks when entering a new market, or raise red flags when releasing new products. We’ve written in the past about the tension between fraud teams and marketing teams, for example.
But what if risk teams could leverage smart strategies to help increase sales and conversions? By turning the sentiment that fraud teams are only about stopping fraud around, fraud managers can actually become heroes. Through cross-functional collaboration, they can identify ways to decrease friction for legitimate shoppers and make the buying experience even smoother.