Will AI Become the New UI in Banking?

Data & Artificial Intelligence

A New User Interface in Banking

Banks are deploying artificial intelligence (AI), or smart machines, to fill in for a range of tasks. Many will replace what humans used to do, while other scenarios will introduce services that were not possible in a pre-digital world. This evolution is leading many bankers to use AI as the preferred user interface (UI) for a growing number of services.  

In this piece from Razorfish, we highlight several ways banks are engaging smart machines, capable of delivering humanlike interactions, to garner an entirely new set of capabilities to attract, acquire and retain customers. Though the application areas for smart machines are vast, we focus on three priority areas:

  • Smart security
  • Smart customer support
  • Smart infrastructure

The trend

In 1959, computer science pioneer Arthur Samuel defined machine learning as a field of study that would give computers the ability to learn without being explicitly programmed. 

Samuel’s early work in artificial intelligence focused on games versus helping marketers deal with the exabytes of data that would be generated by humans and non-humans in a connected economy. Still, he did predict that we’d eventually generate avalanches of data that would overwhelm, even paralyze us as businesspeople. Samuel further predicted that we would look to computers to solve the very problem they would create.  

Today, we’ve entered the age Samuel envisioned—an age in which machines produce insights from data, without being explicitly programmed to do just that. 

Smart Security

A technology becomes smart when it takes humanlike actions based on what it learns from current and past activity. In many cases, the reasoning of a smart machine outperforms that of a human being due to its ability to make more informed decisions as it analyzes hoards of data in real time.  One example of smart machine technology being deployed by many banks today is the smart vision system, designed to curb identity theft while vastly improving user authentication. 

Curb identity theft

Most noted for its use in the automotive industry’s development of the autonomous vehicle, smart vision technology is now heading for a bank near you. 

The technology’s application to preventing identity theft, one of the world’s fastest-growing crimes, led HSBC to install several facial biometric units in its two U.K. data centers to prevent the loss of sensitive information that could lead to a customer and public relations nightmare.  According to Gartner, Inc., the image-processing capabilities of smart vision systems offer the highest performance (and lowest error rates) of all smart-machine technologies. 

Authenticate with laser-beam accuracy 

Improving security and preventing identity theft, always top banking priorities, benefit from smart vision systems that use biometrics to authenticate everything from simple ATM transactions to individual access to buildings (including specific floors and rooms). In the consumer experience arena, many customers of USAA Bank in the U.S. use smart vision’s facial recognition technology to access the bank’s mobile app. For example, a selfie, taken from a mobile phone, is used to approve a transaction (or send a fraud alert). In Taiwan, CTBC Bank’s customers don’t need a card or PIN to withdraw funds; instead, they use facial recognition and fingerprint scanning to authorize common, everyday transactions. Other banks (for example, Wells Fargo in the U.S.) are pioneering smart vision systems to measure consumer attention span in relation to promotions placed in lobbies, airports and waiting rooms.  

The growth in smart vision systems is expected to climb dramatically, aided by the rapid development of driverless vehicles, which will fuel the sale of millions of smart vision systems by 2022. Demand from autonomous car makers will undoubtedly drive prices for these systems down a hundred fold within five years.

Bank Facial Recognition.
In the consumer experience arena, many customers of USAA Bank in the U.S. use smart vision’s facial recognition technology to access the bank’s mobile app. Source: ksat.com

Smart Customer Support

Get intimate with customers

While smart vision systems prevent unlawful access to a customer’s bank accounts, some smart machines are finding their way into customer service to help people get quick answers to common questions. These virtual customer assistants have moved beyond the monotone, robotic sound of old sci-fi films to speak in friendly, humanlike manners. 

Mitsubishi UFJ Financial Group in Japan offers one such example: a virtual customer assistant, fluent in 19 languages, quickly interprets the consumer’s emotional state as he or she is greeted. At the Commercial Bank of Dubai, a virtual customer assistant named Sara is available 24/7 to help website visitors fill out forms and get up-to-the-minute answers to questions about saving and investing. And customers don’t even need a keyboard or a mouse to “talk” to Sara; rather, they wave at or tap on the screen to dive deeper into a topic or initiate a new one. 

Sample vendors in this technology space include EmotionScan, which helps bankers understand how customers feel about their money. Taking a page from the Suze Orman playbook, the software analyzes the facial expressions of customers or prospects as they listen to a series of scenarios designed around topics of cash flow, budgeting, mortgages, retirement, debt and saving. Facial response analysis helps the system interpret emotions, such as a customer’s level of interest or anxiety, to guide the interaction. 

Aldebaran Robotics, part of SoftBank, which markets Nao (a humanoid robot), can implement similar scenarios. Creative Virtual, another provider of smart machine technology, uses its V-Person technology to power “Ask Sara” at the Commercial Bank of Dubai. We expect that virtual customer assistants, designed to handle all common questions and queries, could eventually be deployed as replacements for call center operations.

Offer personal advice

The application of smart machines goes beyond general contact center management to offer virtual personal advice. Personal relationship managers at DBS Bank in Singapore, for example, use IBM’s Watson tool to ensure the quality of the advice they give to private bank customers. Similarly, Standard Bank in South Africa calls on the natural language-processing ability of Watson to gain insights from large amounts of unstructured data.

The bank also uses Watson to speed handle customer queries about their specific situations. Says Vuyo Mpako, head of innovation and channel design for Standard Bank, “The ultimate beneficiaries of the project will be our customers for whom the process—known as ‘cognitive computing’—will undoubtedly bring many benefits as we continue to identify innovative ways of doing business and build a bank for the future.”

The ultimate beneficiaries of the project will be our customers for whom the process—known as ‘cognitive computing’—will undoubtedly bring many benefits as we continue to identify innovative ways of doing business and build a bank for the future.

At Ally Financial in the U.S., Ally Assist, launched with analytics firm Personetics, gives contextualized answers to an individual’s question, submitted by voice or text. Ally also offers predictions and feedback to help individuals manage their cash flows and budgeting, based on real-time analysis of their own spending histories. Ally also uses its knowledge of a customer’s behavior to alert him or her to make a payment. 

We expect the growth of similar scenarios as virtual assistants are deployed at various levels of financial services. For example, we foresee their use in arranging events (e.g., in polling participants to determine the best time and location for a meeting), as well as in job training and professional development. Virtual assistants might even be used to help bank employees identify and connect with work, career or hobby buddies. 

Mobile Banking.

Smart Infrastructure

The internet of things

Smart machines will also be used to lower the cost of building and other infrastructure services, particularly HVAC systems, lighting, people monitoring and security. We expect smart infrastructure to flourish hand in hand with increased adoption of the Internet of Things (IoT). Bank executives with interest in this application of smart machines are starting by collecting more and more data from their physical infrastructures and applying analytics to determine which clusters of reliable signals can be used to predict problems.

As sensing technologies are embedded in building infrastructures (and the assets with which the building interacts), smart machines will help managers make more informed, more reliable decisions about the structural health and maintenance of their assets. This could also transform the approaches architects and developers take to building safety, while leading to the smarter use of green technologies in modern construction.

Regulatory considerations

Federal regulators issue extensive guidance on the deployment of information technology generally; machine learning will prove no exception. The regulatory environment in this area will evolve as providers, working with bankers, find more and more use cases for employing smart machines to deliver better service to bank managers, employees and customers. Emerging regulations will likely use the boundaries already placed on media as the template. 

Rest assured, as Google and other providers turn smartphones into virtual personal assistants, customers will expect banks to interact with them. Hence, banks should not let concerns about regulatory constraints keep them from deploying emerging smart technologies, rather learn to work with them. The best approach for bankers is first to prioritize smart-technology investments and then plan for—and even try to anticipate—evolving regulatory constraints.

Mobile Banking.

Conclusions and Recommendations

Don’t wait

Most banks will invest in smart machines and their supporting technologies over the next few years. Bank CIOs should collaborate with business leaders now to determine how best to leverage these technologies in the areas of security, identity theft prevention, general customer service and personal assistance. Banks with larger appetites and budgets should explore advanced technologies, such as Watson, or the use of natural language-processing technologies, in enhancing decision-making for executives and relationship managers in their private banks. 

Start small

Focus first on setting up virtual and personal customer assistance in a limited set of areas rather than trying to cover your entire portfolio from the outset. Try not to let admiration of the technology lead the way; instead, focus on how (or where, or even whether) customers would benefit from speech-to-text, or how a system that learns from customer interactions would fit into and enhance your existing loyalty program. Always calibrate use cases with industry regulations; for example, offer suggestions tailored to the customer, but don’t recommend financial services products. 

Use smart machines to add value to existing investments

Smart machines don’t necessarily involve a complete overhaul of a function such as customer service. For example, the autonomous vehicle contains several smart machine systems, such as vision and speech-to-text and natural language dialogue. But smart vehicles also contain important, yet unsmart technologies—namely rules engines and sensors.

Monitor adjacent sectors

As the economy becomes more connected, subtle relationships within digital ecosystems become more pronounced. Consider the smart vehicle’s direct impact on financial services as people put greater reliance on Lyft, Uber and Hailo, and as driverless cars turn vehicle usage into a service rather than an individual consumer asset. We are likely to see fewer auto loans being made, along with a dip in individual auto insurance. The risk curve for auto loans, and auto-related insurance businesses, will also be impacted as accidents and injuries involving vehicles fall.

More Insights

Locations

46 Cities.
5 Continents.
1 Vision.