When mathematician Alan Turing and neurologist Grey Walter laid the foundation for Artificial Intelligence in 1943, they never imagined that it would take almost a century to bring their vision to reality. Fast-forward to the post-AI-winter era of today and we are finally witnessing commercial acceptance of the technology, even in traditionally conservative fields such as financial services.
Although banks have never been technologically agnostic, the pace with which the industry has moved from an employee-centric to a tech-centric framework of operations may be an indicator that AI is seen both as an opportunity and a threat.
How India has Adopted Digital Banking
With firm support from the government, India has recently undergone a transformative phase in banking with the onset of demonetization, Digital India, and GST reforms. Although each of these has played a pivotal role in promoting digital transformation, operational gaps still exist. These include:
- Digital Frauds: These present a risk of about US$ 2.8 billion per year to India’s banking system. Hacking, sniffing and spoofing remain serious threats to the industry.
- Authentication Issues: Loopholes in authentication open up possibilities of further threats. Low financial literacy, along with customer awareness, only compound these issues.
- Low Adoption Rates: These are especially prominent among older generations that are more inclined to use conventional banking.
How AI is Changing the Banking Landscape
Artificial Intelligence is enabling banks and financial services to redefine the way they work, create innovative products, and transform customer experiences. Here’s how AI is impacting every vertical in the space:
1. Improves Decision-making for Loans and Credit
AI can provide a more accurate depiction of the potential borrowers at lower costs. By accounting for a number of variables and leveraging relevant data sets, AI is developing new credit systems that can reliably distinguish the difference between high and low default risk applicants. With targeted Machine Learning algorithms and access to personal consumer data (such as smartphone data), the possibility of bad loans can be significantly reduced. In fact, highly trained neural networks can find correlations between thousands of variables, surpassing the capabilities of the human mind by multiple bounds.
2. Reduces Operating Costs and Risks
Operations of traditional banks have been based around human-centric processes and paperwork-intensive workflows. These make tasks error-prone and time-intensive, especially when manual data-handling comes into the picture. With capabilities such as handwriting recognition, combined with natural language processing and RPA, many banking operations can be overhauled. These include managing accounts payable, general ledgers, KYC tasks, account closure processes, and more.
Bank of New York Mellon Corp has already deployed more than 220 bots to execute redundant tasks such as responding to data requests from external auditors and correcting formatting and data mistakes — and the results have been impressive. The fund-transfer bots alone are saving BNY more than US$ 300,000 every year.
3. Facilitates Security Management
AI can seamlessly monitor suspicious behaviors in a digital banking ecosystem by constantly analyzing logs and sifting through fake emails. Additionally, the data associated with such unwarranted activities can be leveraged to prevent and predict security breaches. This is complemented by the analysis of historical data to understand and conduct risk management and extend niche products to customers. Such an intelligent system can also find gaps in compliance frameworks and compare the results with existing bank data for further rectification.
4. Smart Investment Banking and Trading
Since trading has largely become digital, there is ample data that can be leveraged for optimization. The end game of this has been the foray into Machine Learning and AI in smart and systematic investment banking. This is taking shape in the form of automated trading systems that are powered by algorithms to buy and sell orders on the basis of preset conditions. Since these can function significantly faster than humans, they come in handy in micro-volatile markets where the shortcomings of humans are problematic. These shortcomings include inefficiencies such as not following the strategy, error in entering the trade, and emotion-based trading.
5. Aids Regulatory Compliance
With AI, banks can realize regulatory authority and avoid large-scale defaults. Banks in India and around the world have always been subject to strict regulations to fight financial crime. As a result, banks are required to know their customers, prevent money laundering, monitor transactions, and more. AI virtual assistants are used today to comply with all such regulatory needs. JPMorgan, for example, has recently introduced their Contract Intelligence (COIN) platform which can analyze legal documents to extract crucial data points and clauses. Such an AI system can reduce the time taken to execute tasks — 360,000 hours — down to mere seconds.
6. Detects and Reduces Fraud
AI-powered automated fraud detection systems can seamlessly detect anomalies and improve the accuracy of credit card fraud detection. This is done by analyzing variables such as the shopper’s behavior, location, and buying habits to spot events that are out of order. This is essentially how new fraud detection platforms (such as Plaid) work.
The strategic investment of Citibank in Feedzai reiterates the importance of AI in the banking ecosystem. Feedzai is a data science platform that functions in real time to identify fraud in both digital and in-person banking with the help of large-scale data analysis.
7. Enhances Customer Support and Experience
Since AI systems can simultaneously churn multiple data points, they can also quickly learn about the needs and preferences of customers. Further analysis of behavioral patterns enables banks to combine information from various channels such as transactional histories, queries, searches, and even social media accounts. As a result, they can extend personalized products to every customer, significantly boosting retention. And that’s not all — virtual assistants are improving customer experience in banking at every touchpoint. For instance, Bank of America has introduced a smart virtual assistant named Erica that can leverage predictive analytics and cognitive messaging to extend automated financial advisory services to more than 45 million banking customers.
As banks prioritize strategic technological advancements, AI is taking centerstage. Banks are increasingly finding use cases for AI in verticals such as security and business intelligence (at the banking end) and conversational interfaces (at the customer end). Although the approach differs from one bank to another, one thing is certain — the AI revolution in banking is here to stay.