The FinTech startup ecosystem infused about $14.5 Bn globally just in venture funding in 2015 which is almost twice from what it gathered in 2014 and next year, an astounding amount of $36 Bn. i.e which is an increase of 5 x in 2 years and perhaps rightly so because it impacts everyone. This has allowed innovation and flexibility in financial services. With the cross domain application capabilities of big data and machine learning, it has in itself surrounded the main business model of plenty such startups in FinTech especially in the cyber-security domain. In the coming years it is expected to play a major role in the same with its application expanding in Predictive Analysis for Credit Scores, Fraud Detection, Identity Management, Trading Algorithms and Information Extraction etc.
However, advances in machine learning to ensure cyber-security in FinTech also makes it vulnerable to attacks that are based on Machine Learning algorithms. Not only that, it would make it hard for the system’s human designers to understand system behavior and it’s associated security weaknesses.
This scenario resembles an arms race where machine learning has made its entry and increased the complications. This puts the system designers in a spot when building machine-learning based platforms as they now need to pay particular attention to the biases that could be inadvertently built into solutions and thus create cyber-security vulnerabilities by themselves.