Gurucul sponsored the Money20/20 show in Las Vegas last month. We focused on two tracks: AI & Deep Learning and Cybersecurity & Fraud. Below are some of the key takeaways.
Artificial Intelligence Is Embedded in Fintech
Applying the technology to the financial services industry has been transformational. It’s clear that AI is already embedded in Fintech. Kartik Gada, Executive Director of Woodside Capital Partners commented, “AI is like air – we would only notice it if it was gone.”
Greg Michaelson, General Manager of Banking for DataRobot made a very interesting point. Even though AI is automating and making businesses smarter,they’re still doing really stupid things. You make a call to your cell phone provider. What do they say? “Enter your phone number, followed by the pound sign.” These are the people who invented caller ID. They know everything about you: where you live, what emails you get, who’s texting you, etc. Why can’t they predict why you’re calling? Why are we spending billions of dollars on IVR technology when we could be making great predictions? Your mobile provider sends you a notice for not paying your bill and you call them. It’s pretty obvious why you’recalling, but they still put you through the system. Greg’s point: “There’s smart AI and there’s dumb AI and there’s certainly decisions about where we should be innovating.”
Our point: make sure you choose a partner who can help you focus your investments on smart AI implementations. However, it’s easy to get this wrong.
History is littered with companies who didn’t embrace change. As Tom Poole, SVP Digital Payments & Identity at Capital One points out: Kodak invented the digital camera, but it put them out of business. Similarly, Netflix tried to sell themselves to Blockbuster, but they wouldn’t buy. Then, Netflix put Blockbuster out of business. If large incumbents get enamored with their existing process and stop looking at how the market shifts, they can get left behind by an entirely new approach to their ecosystem. Certainly, you need to embrace AI and ML, or you will be left behind.
Define Your Roadmap
What should financial companies be doing with AI and Machine Learning? Companies ask, “What are others doing in this space?” They are all trying to get to the use case. There are literally thousands of AI and ML use cases in Financial Services. The right question they should be asking is, “What does my roadmap look like?” Business people need to learn the tools to spot opportunities, and have the technologies available on hand to actually build the solutions. Every organization needs a roadmap with good number of items on it that they are going to build in the next 2 to 5 years.Then, they need to figure out the tools they need to get there. We have you covered.
Your Biggest Challenge is Fraud
Fraud is a massive problem: card not present, ID theft, synthetic ID, social engineering, phishing, real-time payment fraud, call center fraud, ATM fraud – the list goes on and on. There are twice the number of new social security numbers. Which ones are good? Which ones are bad? SSA has no funding for ID verification.
Jim Hickman, AVP Financial Crimes Operations, USAA acknowledged, “We are all fighting the same fraudsters. Nobody has a competitive advantage.” In other words, the biggest competition a bank has is fraud, not other banking institutions. Fraud threats have escalated from thousands per hour to millions per hour. The only thing you can do is reduce risk.
Richness of Data is Key to Identifying Fraud
Banks have a lot of data, but it’s really hard to get that data out. It’s super siloed. Often, the lines of businesses have different databases. Even if you can get the data, is it in the correct format? Do you need to duplicate it? It’s really important data you need to analyze, but it’s a major task to turn that data into information. Analytics makes sense of data. So, our analytics ingests all the various data sources, correlates that data, does link analysis, and discovers hidden patterns that can help organizations make informed, risk-based business decisions.
Banks haven’t connected the dots across customer touch point data. Many things individually do not look suspicious, but if you look at them collectively you can identify patterns. Machine learning on big data enables you to find patterns in massive amounts of data, so you can connect the dots.
Gurucul Fraud Analytics
Gurucul is working with companies across many industries to address their fraud detection and prevention needs. While there are many different use cases, the theme that is common among them is that organizations want the ability to do cross-channel fraud detection, to aggregate and link more data coming from many different systems. It is this cross-channel capability that shines a brighter light on not just transactions but also subtle behavioral activities and peer group analysis that would otherwise go undetected.
In conclusion, legacy Fraud Detection solutions cannot curate the amount of data across all the channels needed to connect the dots. Gurucul Fraud Analytics provides a holistic risk-based approach for fraud detection that works in tandem with legacy solutions. In many cases, the fraud can be detected in real time such that action can be taken to prevent loss from the fraudulent activity.