Do you remember Deep Blue, the supercomputer that defeated the legendary chess player Garry Kasparov in February 1997? That epic contest between human and artificial intelligence marked the first victory of a computer against a reigning champion during a tournament. And it certainly provoked many more conversations in academia, industry, and elsewhere about the forthcoming rise of the machines and how technology could one day take over humans.
Well, 21 years down the line, artificial intelligence (AI) is finally moving beyond the realms of research and sci-fi and making its presence felt in the real world amid an unprecedented convergence. Advancements in high-density parallel processing infrastructure and an extraordinary surge in the volume and type of data being generated are fuelling growing adoption of machine learning and other cognitive technologies. Also enabling this trend is the increasing adoption of cloud computing and mobility as well as open sourcing of machine learning (ML) algorithm writing.
The Business Imperative
Like other industries, the financial services sector, too, has started using AI and related disruptive technologies, albeit on an incremental basis. Research and Markets expects the market for AI software in financial services to expand at a compounded annual growth rate of 40.4%, from $1.3bn in 2017 to $7.4bn in 2022.
Market for AI software in financial services will expand at a compound annual growth rate of 40.4%
The business case for financial institutions (FIs) to leverage AI is pretty compelling. As ultra-low interest rates continue to impact their profitability, banks are eyeing opportunities to boost operational efficiency by reducing unscheduled downtime of systems, and “rightsizing” headcount based on evolving demand. Growing the top line through enhanced personalized targeting of offers and optimization of sales processes is also emerging as a focus area for FIs. The third major driver of artificial intelligence adoption across the financial services industry is the pressing need to comply with increasingly heavy regulations across different jurisdictions. Amid introduction of norms including the European Union’s Payments Services Directive (PSD2) and the U.K.’s Open Banking program banks now want to harness rich customer data for building better apps and services. Finally, finance is realizing the significant potential of enhancing marketing effectiveness and customer service via intelligent automation.
According to PwC’s 2017 Digital IQ Survey, almost 52% of respondents belonging to the financial services sector cited ongoing “substantial investments” in AI with 66% projecting significant investments by 2020. And 72% of the senior management sees AI, ML, and other cognitive technologies as a key source of competitive advantage in the near future, the survey showed.
Emerging use cases
So, what are some of the notable emerging applications of artificial intelligence and machine learning in the financial sector? While banks, asset managers, and insurers have rolled out several pilots across the front, middle, and back offices, let’s look at the following key initiatives:
Credit and Insurance Underwriting: Lenders and insurers are using machine learning to process credit lending and insurance applications faster, and in a cost-effective manner, without diluting risk assessment standards. The key here is accuracy, scale, and speed with machine intelligence clearly proving to be far superior to humans in analyzing massive volumes of consumer data. For instance, FIs have begun using self-learning algorithms (SLAs) to mine millions of consumer data sets concerning age, job, marital status, credit history, and so on in order to establish the risk profile of individual applicants.
Fraud Detection: Unlike conventional financial fraud discovery systems that were based primarily on a predefined checklist of risk factors and a complex set of rules, ML-driven fraud detection proactively spots “anomalies” and flags the same for security teams. The ability of intelligent algorithms to proactively anticipate potential fraud with an infinitely larger data set is likely to reduce false positives, instances where “risks” are flagged that turn out to be a false alarm.
Workflow Automation: FIs are already using natural language processing (NLP) to automate certain business processes in an effort to rationalize expenditure and boost customer satisfaction. Machine-based interactive customer service in the form of chatbots is gaining traction. Bank of America’s intelligent virtual assistant, named Erica, harnesses predictive analytics and cognitive messaging to perform “day-to-day transactions” beside analyzing the unique financial requirements of each client and recommending relevant solutions.
JPMorgan Chase’s Contract Intelligence (COiN) platform, meanwhile, deploys image recognition software to review contracts and other legal documents in seconds, as compared to the 3,60,000 hours it takes to manually scrutinize 12,000 annual commercial credit agreements.
The Bank of New York Mellon, the giant custody bank, has implemented “bots,” robotic process automation (RPA), in particular, to enhance operational efficiency. The firm says RPA has enabled 100% accuracy in account-closure validations across five systems, and improve processing time by 88%.
Asset Management: So-called “robo-advisors,” such as Betterment, are offering algorithm-based, automated financial planning solutions to clients, helping them build investment portfolios that are aligned with their individual goals and risk tolerance. Wealthfront leverages AI to track account activity of each customer to mine account holders’ spending and investing patterns, in order to be able to provide bespoke services.
Algorithmic Trading: Many hedge funds and other “buy-side” participants in financial markets are using complex artificial intelligence-systems to make thousands or millions of trades in a day. These systems, based on machine learning and deep learning, are facilitating “high-frequency trading” (HFT) by analyzing a wide range of market factors in real time.
Conclusion
As the pace of technological innovation breaks down barriers, the financial services industry needs to prioritize its goals and develop long-term strategies. We are already seeing signs of a certain maturity among CIOs as they weigh in on their investments. Recent tech developments, such as automated assistants that Google showcased in their recent I/O and intelligent image recognition, have great potential for financial services and firms are now deciding their investments with an eye toward capturing the next generation of digital-native customers.
While the adoption of artificial intelligence and other cognitive technologies will only accelerate across the financial services sector, FIs will have to keep a couple of things in mind in order to maximize ROI.
- One, the quality, and reliability of data they gather and process must be beyond doubt. The failure to feed in top-quality data will result in poor outcomes, as far as AI and machine learning-based analytics initiatives are concerned.
- Second, financial firms will have to factor in the need for robust data security, while designing and implementing machine intelligence programs.
The journey ahead will not be easy, but banks and insurers must keep experimenting and iterating if they are to remain relevant in an increasingly uncertain world.
