You are here
Home > Tech > Tech Talk > Conversational AI for Banks: What’s the concern?

Conversational AI for Banks: What’s the concern?

Bank of America’s virtual financial assistant Erica is cited as a prime example of how financial institutions can introduce AI concepts to everyday consumers through digital banking experiences. Large banks and financial companies have also started to offer banking through virtual assistants — Amazon’s Alexa, Apple’s Siri, and Google’s Assistant — in a way that will allow customers to check their balances, pay bills and, in the near future, send money just with their voice.

In fact, large banks are leading the AI revolution in financial services – nearly half (48%) of banks with over $50 billion in assets have already deployed AI solutions, compared to just 7% of banks with $1-10 billion in assets. However, ensuring the bank is ready for conversational AI is a major consideration according to Dror Oren, Chief Product Officer & Co-Founder, Kasisto.

Access to real-time data is crucial for creating personalized and contextual conversations with the virtual assistant or bot. In a podcast interview with Karen Webster, Dr. Akli Adjaoute, CEO of Brighterion, a Mastercard company, pointed out that the legacy machine learning and AI currently powering digital assistants like Alexa, Siri and Cortana lack both adaptability and the ability to determine context.

Different use cases require different data sets, explains Dror Oren. For instance, a seemingly simple question like, “How much money do I have?” requires coordination across all bank accounts, not just a checking account. On the other hand, a query like “How much did I spend on my trip to New York?” requires data to be augmented with geographical information.

Dror Oren feels that creating content for entirely new customer experience is crucial. Content needs to be shaped and refined for conversational interactions. That means short, bite-size replies – paragraph-long answers that are appropriate on your website are out of place in a conversation with a virtual assistant. “The only way you can have a really intelligent assistant is if the virtual assistant is able to understand something without any pre-programmed language,” said Adjaoute.

The potential to do sensitive tasks like sending money to friends and family instantly with voice commands through a smart speaker raises security concerns. Virtual assistants and smart speakers are still relatively new technologies, and potentially susceptible to being exploited by cyber criminals. “Users’ voices can be recorded, manipulated, and replayed to the assistants,” said Kurt Baumgartner, a security researcher with Kaspersky Lab. “Also, with access to banking accounts and abilities to transfer and pay out money, remote financial fraud may be within the reach of cybercriminal groups soon.”

It is critical for security, compliance and legal authorities to believe and support the idea of virtual assistant because without involving these key stakeholders in the process, securing cross-functional collaboration and approval is impossible according to Dror Oren. Educating these key stakeholders early about the inherent differences of a conversational AI experience will help the bank avoid a potential delay-inducing stalemate down the road.

Deployment is another key consideration. While banks are early adopters of AI solutions, they lag further behind in cloud adoption due in part to security concerns, explains Dror Oren. Confining conversational AI to solely on-premise environments carries some disadvantages. One of the inadequacies is that banks do not benefit from instant AI training improvements as a result of continuously learning from customer interactions.

Technology expert and syndicated radio host Kim Komando said users of voice-activated devices should be aware that everything they say could be recorded. “You always have to remember that when you have these devices in your homes … they are truly always listening,” Komando told KTAR News 92.3 FM. “The other thing that you have to remember is that anything that’s ever connected to the internet – and I know it’s frightening – but it is truly hackable,” said Komando.

Ajit Chawla, vice president of Birlasoft, believes it’s still early days for how banks will weave AI into the customer experience. “In the next two to three years, we’ll see a lot of mainstream adoption of AI,” he said. “Right now, we’re still at the beginning stage. Eventually, AI will help bring some standardization to user experiences.”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.