AI in the banking industry is not a novelty. In fact, as early as 1950 – with the appearance of ATM’s – check automation and recognition technology, early forms of AI, started becoming available to bank’s clients.
Since then, AI technology has advanced significantly and given way to what some experts call the ‘Fourth Industrial Revolution’. Judging by internet search trends, investor interest and the amount of startups launching products that incorporate AI, at a fast clip, one could assume we are in the ‘rise-of-AI-era’. What does this mean for FinTech and how much of this wave is potential versus reality? How will it translate to the financial products clients consume? Is AI the engine of core business in FinTech?
What does AI look like in FinTech?
Despite all of the attention, the overwhelming majority of innovations fall into one of two buckets. First, improving customer experience, through identification and verification products which provide more automation in the application of insurance, loans and credit lines, as well as, the use of chat-bots or virtual assistants in financial customer service departments. Second, reducing costs, when AI is used to detect fraudulent behavior, suspicious transactions and prevention of cyber-security threats.
FinTechs that use AI as their central value proposition have yet to prove their value add, although in the consumer and SME lending space, there are some seed use cases with promising Startups. Affirm is aiming to overhaul existing credit score systems like FICO and open up the previously untouched, non-prime consumer section of the market using AI. A radically different approach to traditional finance where models rely on cost shifting and backend pricing schemes to pay for poor underwriting and attempt to increase profits at the expense of customers. Affirms’ CEO, Max Levchin, said his motivation is “to provide a transparent alternative that doesn’t hijack human psychological blind spots”.
Another example is ZestFinance, which claims to have solved this issue with a new credit-scoring platform, called ZAML. The startup sells the machine-learning software to lenders and also offers consulting services. Other startups in the space include C2FO, FundBox or Numer.ai. What is most striking is that incumbents or more established FinTechs in lending are not using AI for their processes but instead rely on statistical models, third party verification and the human element. When will this gap close? Where do these two models converge?
Why is FinTech different from other industries?
Incumbents and other firms that don’t rely on AI attribute it to several reasons: regulatory and compliance barriers, challenges in the organization’s data architecture, low and slow innovation penetration resulting from a lack of dedicated teams to test new ideas and concepts; and budgetary constraints. Surveys show that as many as 12% of incumbents considered AI “new, untested and risky”.
Beyond these reasons, other barriers are rooted in the human factor and on whether banks and FinTechs understand their customer on a human level. When it comes to financial services, this relationship is characterized by a lack of trust, which is a legacy effect of the financial crisis. For example, in the UK alone, less than a third of adults feel financial firms treat them honestly and transparently. Banks and FinTechs should be on a mission to fix peoples broken relationship with money. While banks and FinTechs move towards more reliance on technologies like AI, does this necessarily mean less appeal in person? Across FinTech we find plenty of user-friendly interfaces and other tools to bring customers more “human” experiences, but are they also better tailored and more personalized to their customers? Perhaps this is what we are not fully capturing.
In the coming years we will likely see more tangible performance for AI products in FinTech, especially as banks surpass the above mentioned challenges and early-stage startups are able to raise more funding. Areas that are expected to thrive in FinTech are those that touch upon the consumer relationship in financial services and improvement current state-of-the-art machine learning in cyber-security, payment intelligence and info-security intelligence.