It is sometimes scary thinking about all the data companies collect on us – it can feel like an invasion of privacy. Why do companies need to know what we’re buying, where we’re going and what ads we click on?
Now while there is a case for too much data sharing, there are also many benefits to companies holding our data. One such reason, is fraud detection. Fraud is a big problem for Canadians as in 2017 over $95 million was lost from scammers. From this $95 million lost, $13 million was due to e-commerce scams. As more and more of our shopping happens online this number will likely rise, unless we do something about it. One of the most effective ways companies have been combatting fraud in recent years, is through data collection, AI and machine learning.
AI & Machine Learning
AI and machine learning seem to be the biggest buzz words in tech right now – but what exactly do they mean?
How many times have you used Siri or browsed through Amazon’s recommended purchases? Odds are, very often. Not only do these actions make your life a little easier, they also show that you have used artificial intelligence (AI). At the most basic level AI describes anything a machine does that requires some kind of “intelligence”. In this context, intelligence can be analyzing, thinking, problem solving, reasoning etc.
Imagine if you came home from work to find your computer reading a textbook about detecting fraud. You’d probably run away screaming, thinking you were in some weird new sci-fi movie. But is the idea of machines learning really that crazy? Not anymore. Although computers don’t learn through reading a textbook, they do learn and have been for many years. This is called “machine learning” which is a form of AI and refers to anything computers “learn” how to do without being programmed to do so. Usually, machines “learn” through going through large amounts of data and finding patterns, trends and characteristics that they then use to “learn” how to do something else. The more data you feed your computer, the “smarter” it can be.
For years banks have been using machine learning to detect fraud faster and more efficiently. By feeding them data about consumers spending behaviours, computers learn different trends and links in our spending patterns. Computers will begin to understand what normal spending is like for our everyday lives, for when we go on a vacation or for when we are buying a new home. Based on these “normal” behaviours, computers can then learn how to spot anything that doesn’t fit and thereby find fraud.
A major benefit to using computers and machine learning, is that computers can sort through large amounts of data much faster than humans can. This allows for fraud to be detected sooner and your money to be safer. Additionally, unlike humans, computers don’t have biases and they don’t get tired. This allows for computer analysis to often be more accurate and consistent in detecting fraud.
Humans are not perfect, and either are computers. Although computers beat humans when it comes to quickly sorting through data and removing bias, they have their shortfalls. Computers cannot give solutions and may miss fraud if they haven’t “learned” a certain tactic yet. Therefore, the best strategy is to have analysts and computers work as a team to detect fraud and protect consumers.
We’d love to chat! For all things digital reach out to email@example.com
Horan, T., & Analytics & Optimization. (2018, July 31). 5 Keys to Using AI and Machine Learning in Fraud Detection. Retrieved from https://www.fico.com/blogs/analytics-optimization/5-keys-to-using-ai-and-machine-learning-in-fraud-detection/
Council, Y. E. (2018, June 05). Artificial Intelligence And The Future Of Financial Fraud Detection. Retrieved from https://www.forbes.com/sites/theyec/2018/06/04/artificial-intelligence-and-the-future-of-financial-fraud-detection/#689ef7f2127a
Reitblat, M. (n.d.). Artificial Intelligence, Fraud Prevention & Human Intelligence. Retrieved from https://blog.forter.com/artificial-intelligence-fraud-prevention-next-gen-machine-learning
Fraud detection and machine learning: What you need to know. (n.d.). Retrieved from https://www.sas.com/en_ca/insights/articles/risk-fraud/fraud-detection-machine-learning.html
What is Artificial Intelligence (AI)? – Definition from Techopedia. (n.d.). Retrieved from https://www.techopedia.com/definition/190/artificial-intelligence-ai
Heath, N. (2019, April 01). What is AI? Everything you need to know about Artificial Intelligence. Retrieved from https://www.zdnet.com/article/what-is-ai-everything-you-need-to-know-about-artificial-intelligence/
Caswelllogan. (2018, October 31). Over $95M lost because of scams in 2017: Better Business Bureau. Retrieved from https://globalnews.ca/news/4052845/over-95-million-dollars-lost-because-of-scams-in-2017-better-business-bureau/