- 16th April 2018
- Posted by: criticalfuture83
Brian O’Donnell is executive in residence at the Global Risk Institute in Financial Services.
“I don’t trust banks. I believe when the robots rise up, ATMs will lead the charge.”
– Sheldon Cooper, The Big Bang Theory
Bank customers can be forgiven for wondering how Facebook and Google can seamlessly anticipate and fulfill their requirements, while their bank of 30 years cannot do the same. After all, banks have far richer data about us than any social media site, yet they have not been able to use that information to efficiently predict our future financial needs.
This shortcoming can be attributed to a number of factors including legacy systems, complex compliance requirements, old-school cultures and an understandably cautious approach to new technologies. But that is about to change.
Taking their cue from social media, banks are now embracing a host of new technologies such as artificial intelligence (AI) – including machine-learning algorithms and natural language processing – application-based services, cloud storage and real-time data management. A number of factors are driving this innovation, including the migration of clients to digital channels, technology advancement in data management and analytics, and ever-rising expectations of clients and regulators.
Social media pioneered cloud computing and developed data management software out of necessity to manage costs and develop valuable insights about clients such as what they like to buy, do and watch.
As Facebook, Google and others launched apps to serve and learn about their users, they employed cloud storage and AI to drive down the cost of storing and managing data. This allowed them to cost-effectively present meaningful results to users, and actionable results to advertisers so they could more effectively target the right audience.
This scenario is supported by the findings of a recent paper published by the Global Risk Institute (GRI), which explores the benefits of AI for the financial services industry. The paper’s authors note that when conditions are ripe – that is, when data, technology and a suitable need converge – there are significant opportunities to use AI to analyze problems in more efficient and effective ways. They do not predict the wholesale replacement of humans with machines, but rather a redirection of human resources to more value-added sales and service roles.
So how will all this change the world of banking? Luckily for banks, there is no end to potential uses.
By leveraging AI, banks can engage with consumers in a faster and more consistent manner. They can use “bots” at contact centres for basic inquiries to free up employees for more complicated questions. They can use robo-advisers to provide basic investment services at lower cost.
They can put digital payment adviser applications on mobile devices that will advise consumers on which credit or debit card will earn the most points when they make a purchase. And they can upgrade their cybersecurity and fraud prevention with biometric technologies such as voice recognition. In fact, the potential for banks to boost earnings and save money with AI and related technologies will skyrocket in the next five years.
The challenge for financial service providers, like all industries with legacy technology, is finding ways to dovetail their valuable data together with modern systems and networks. Legacy client applications and task-specific data warehouses mean that, unlike the social media industry, banks don’t naturally have all of their client and transaction data available for analysis, the way that the newer social media companies do.
For now, banks continue to suffer from the 80/20 rule – their data science experts typically spend 80 per cent of their time searching for, correcting and consolidating data and only 20 per cent actually developing algorithms for analytical insights. But as banks migrate to the big data tools developed by the social media industry, they will be able to flip the 80/20 rule on its head and deliver better business insights for the business and their clients.
Anti money laundering (AML) is a good example. Regulators require that banks monitor all transactions for anything out of the ordinary – say a large cash deposit or a cash transfer to a far-away tax haven. A case worker must spend endless hours scrutinizing and reporting all suspicious transactions. With existing technology architecture, this is often a very expensive and inefficient process, resulting in thousands of “false positive” alerts – which still must be closed to the regulator’s satisfaction.
However, a predictive intelligence software company called Tresata has developed technology that can greatly streamline AML processes. In a recent paper for the GRI, the U.S.-based firm shows how a combination of big data tools, advanced analytics and machine learning can improve the accuracy of AML reporting. The result is the holy grail of anti-money-laundering management – identifying the actual activities of fraudsters and terrorists, without the endless noise and inefficiency of false positives. These tools will likely be adapted to other routine tasks such as fraud and credit risk management.
So, sorry Mr. Cooper, the robots are indeed starting to rise up, not as a threat but as a defender of our banking system. Indeed, the industry needs to readily engage AI, cloud computing and advanced data management technologies to remain relevant to consumers.