Co-authored with Ramtin Matin, Lead Technological Strategist, SpareBank 1 SR-Bank
Artificial intelligence is one of the key components behind the digital revolution in the banking and finance sector. From fraud protection to customer service, there is a seismic shift in how these companies are using technology—and AI specifically—to improve
product portfolios and bolster bottom lines.
One popular aspect of banking which is ripe for disruption is financial planning. Previously thought of as the exclusive purview of upper-income earners and high net-worth individuals, advances in conversational AI are democratizing this space and making
it possible for everyday consumers to get certified financial advice in an entirely new way—via automated online chat.
Meeting the regulatory challenge
Unlike other core banking functions that are already handled with conversational AI-powered virtual agents, such as funds transfers and credit card blocking, the path to automated financial advice is a little more complicated. This is in large part due to
strict regulations, and collaboration with financial authorities and regulators themselves is crucial to bringing this game-changing technology to bear.
report published by the Financial Stability Board about the cross-section of AI and financial services, a number of benefits and risks were identified. Chief among them is the lack of interpretability or auditability of AI, meaning machine learning methods
could become a macro-level risk without appropriate oversight. Just as a human advisor needs to be suitably qualified in order to make investment recommendations, so too the digital processes of a virtual agent must be certified to ensure they meet the same
stringent guidelines. This is the only way to ensure all possible risks to the customer are taken into consideration.
The Norwegian financial authorities, for instance, are currently on track to
roll out financial planning certification by March of this year, which will require adherence to a strict set of standards. Sparebank 1 SR-Bank, one of the largest savings banks in Scandinavia, is currently in the process of upgrading its existing virtual
agent called Banki, with the necessary functions and content in time to participate in this certification structure. The goal is to become the world’s first officially certified digital financial advisor.
Sparebank 1 SR-Bank is widely viewed as a pioneer in the use of conversational AI in banking customer support, with Banki credited with expanding the firm’s digital capabilities, while reducing the costs associated with its growing customer service needs.
The results have been widely championed, with the virtual agent automating more than three-quarters of all B2C and 39 percent of all B2B customer support traffic. It is now effectively handling the equivalent work of 31 full-time employees, and has increased
support capacity by an astounding 175 percent without hiring any new staff.
The first certified digital financial advisor
In developing this advanced functionality for Banki, or any other digital financial advisor for that matter, the biggest challenge lies in risk assessment. There are plenty of bots on the market today that serve as robo-advisors to help people buy stocks.
While they are capable of doing simple mathematical calculations, certified financial advice is far more nuanced.
When a customer sits with a financial advisor face-to-face to determine a retirement plan, for example, the calculations are relatively straightforward. However, the accompanying conversation around risk-aversion, unforeseen expenses and other less easily
explained challenges, are equally as important in helping the advisor make accurate recommendations.
Leaving out this critical aspect of the transaction, as uncertified advice chatbots typically do, is unethical and could be detrimental to a customer’s finances. Not to mention, it could also open up a bank to potential litigation.
The advantage of conversational AI is that, utilizing advanced natural language technologies and deep learning, a virtual agent can be trained—within specific guidelines set out by the financial authorities—to conduct the same ethical and objective risk
assessment that a human advisor would.
Building trust is key to success
Once the hurdle to achieving human-machine advice parity has been cleared, the potential for what a virtual financial advisor can bring to the banking sector is enormous.
In a recent survey, conducted by third-party research organization
SINTEF, it was found that conversational AI is truly an age-agnostic technology. Customers from as young as 18 up to the age of 70 or more, are regularly interacting with Banki, and three out of every four of them choose to get help from the virtual agent
even when given the option to talk to a human.
This indicates that banking customers are not only savvy consumers of digital channels, but that they increasingly want better self-service options to manage their finances. This makes sense. When signing up for a first mortgage, most consumers want assurances
that their finances can handle the added expense. But those who already have experience with purchasing property may prefer to expedite the process in a quick and efficient manner. This is where a virtual agent can really help as it plays to those strengths
while also remaining objective.
Moreover, when trained correctly, an artificially intelligent financial advisor can be inherently less biased than its human counterparts, avoiding prejudices towards ethnicity, gender, religion, for example, that a human might even subconsciously allow
to influence their decision. This is a powerful tool for satisfying diversity goals and meeting non-discrimination mandates.
A decade or more ago, consumers could not have anticipated signing up for a home loan on their mobile phone, but today it is commonplace. In the next 2-3 years, we can expect to see customers handling nearly all of their mortgage advisory process, retirement
planning, or even their investment portfolio, via a virtual agent.
As the technology powering conversational AI continues to develop, and the industry finds new and interesting ways to take advantage of it, the resulting interactions between customers and machines can be a powerful vehicle in building trust towards a bank
or financial institution. And, as we know, trust is the highest form of currency that any brand, not just a bank, trades on.