Artificial intelligence (AI), machine learning, deep learning, cognitive use of AI, and hybrid learning models – these have all lately been at the core of operations or conversations surrounding most successful fintechs and even traditional banks, reeling under the effects of legacy infrastructure.
Some pertinent questions that noted observers in the industry have raised over the years are:
- Is AI overhyped?
- Does the industry have unrealistic expectations from it?
- Is the fintech industry and the financial services sector adopting it in the right way?
According to the Fintech in 2018 report published by FleishmanHillard, some experts have predicted that eventually AI may become more dominant in the back-end processes that improve operational efficiencies rather than customer-facing processes and operations.
While it may be too early to predict the exact nature of the role AI will play over the next decade or so, here are the key areas where AI has already acted as a disruptive force:
- Deep neural network
- Capsule neural network
- Deep learning
- Probabilistic programming
- Automated machine learning
- Explainable artificial intelligence
Companies at the forefront of AI innovation have used these technologies to transform the banking topography through a stronger synergy with the traditional banks, leading to increased profitability, customer-centricity, and financial inclusion.
China and Singapore Seen Pushing the AI Envelope in Asia
According to the KPMG Pulse of Fintech Report, China accounted for more than half the total investments in AI that happened in Asia in the third quarter of 2017. China invested US$745 million, which helped the continent push past the US$1 billion mark for the first time.
Singapore too continued to play an active role as a foremost creator and adopter of AI in 2017, witnessing deals worth over US$25.3 million in Q3 2017.
The government of Singapore and the Monetary Authority of Singapore (MAS) are also encouraging innovation and use of AI in fintech. Digitisation of financial services is key, and integration of services between banks and Fintech companies like Bankbazaar is an important way of taking this forward.
Last year, they announced a grant worth S$27 million for the industry as a part of the Artificial Intelligence and Data Analytics (AIDA) grant. The grant would be offered under two tracks – the research track and the financial institution track.
Under the research track, the aim is to:
- Co-fund up to 70% of the AI and data analytics projects of research institutions that have a direct impact on the financial services industry.
- To promote industry sharing and local knowledge transfer.
- To facilitate right skilling of employees who’re being redeployed into newer roles.
Under the financial institution track, the primary aim is to:
- Co-fund up to 50% of costs for fintech companies in Singapore working on AI projects that help glean better customer insights and improve predictive analytics.
- Improve customer experience and user interface through creation of powerful interactive technologies, machine learning, and natural language processing, among others.
Things That AI Has Achieved So Far:
- Increased security: AI has increased the processing speed of machines at play. Therefore, it can mine large quantities of data and identify certain patterns that lead to thwarting of future attacks on systems and immediate identification of fraudulent transactions. It can also automatically verify documents necessary for conversions like account creation, loan consumption, and origination. This creates a greater level of automation and also increases the processing speed of systems supporting customer education and conversion.
- Better customer support: With more intelligent chatbots and AI functions being infused into the customer-facing services, the level of customer centricity is improving. With AI, there is little or no downtime. Further, it continuously studies the behaviour of a customer and provides interesting insights that help a business improve its level of customisation, detailing, and accuracy while offering products or services.
- More accurate appropriation of company budgets: Whether a company is trying to draw up its marketing budget or understanding how much needs to be spent on customer-specific research, AI helps them get a more accurate estimate of their specific cost involvements and resource allocations. This is crucial to operational efficiency and profitability of fintech players.
- Better credit profiling: Assessment of credit-worthiness of a customer based on simple parameters used by credit bureaus is no longer viable. Alternative lenders are depending on alternative data like e-commerce purchases, hyper-local activities, social media profiles and interactions, and browsing data, to create a more reliable credit profile of a customer. To access and mine these massive volumes of data, these institutions are using AI-based algorithms and software integrations.
- Reducing human error: Certain customer-facing processes like documentation and verification are more prone to human errors because it involves a lot of cognitive intelligence and brainwork. By introducing AI in verification processes and data analysis, companies are not only reducing human error but also reducing duplicate expenses.
Things That AI Can Achieve Going Forward:
- Explainable predictions: Simply predicting a future customer action based on results from the past may not be reliable. Instead, if AI can understand the cognitive/logical path of reasoning that led to a certain decision, predictions into a customer’s future decision-making will become even more exact and error-free. PwC has predicted that it may be adopted by most fintech companies in the near future either as a best practice or due to regulatory needs.
- RegTech: A largely complex and non-transparent regulatory framework means that large volumes of data have to be computed, processed, and assimilated before critical decisions can be made. While this will eventually lead to open banking and greater transparency, the challenge lies in educating the customers as necessary and removing their pain points.
- Emotion analysis: Behavioural insights gleaned are meaningless without understanding their emotional impact. AI, in the future, may be able to suggest songs simply based on your mood patterns. Emotional AI will largely depend on human psychology, facial expressions, voice modulations, changes in brain waves, etc. to understand what you exactly want and suggest products or services accordingly.
- Real–time risk assessment: Real-time updates are extremely important for the financial services sector. The latest AI technology will be able to turn real-time updates into instantaneous business insights – opportunities and threats, based on which businesses and their clients will be able to make effective decisions.
That AI can generate actionable insights for businesses and their customers is a foregone conclusion. The question remains whether it will reach the level of maturity that is expected from it and create new synergistic opportunities for the industry and levels of customer value that have never been experienced before.