Artificial Intelligence (AI) is set to become mainstream technology for customer support for financial services. This was one of the highlights and focus of the recent Digital Banking conference organized by American Banker in Austin TX in June.
Currently, US banks are using chat interactions via Instant Messaging (IM) solutions such as Facebook Messenger or Twitter. Human2Human (H2H) chat is quickly being replaced by Human2bot software capable of managing thousands of customer scenarios, as defined by Q&A templates.
Customers can ask a question to one of these chatbots via chat messages and receive (theoretically instantaneously) a text or voice answer based on AI analysis of the question’s meaning, and an ‘understanding’ of its pre-defined template answers. And for the most critical queries that automated heuristics cannot yet tackle, humans can drop in to fill any gaps.
AI chatbots are starting to take the place of traditional customer care hotlines through the use of Interactive Voice Response (IVR) technology. Engineers feed in thousands of scenarios that represent typical customer queries and their meanings into an AI, enabling it to interpret queries.
Currently IVR solutions use H2H interactions, relying entirely on IVR personnel training and scenario databases to empower their staff to address the most common queries, and flag aggressive and unusual ones for escalation.
The benefit of AI knowledge hubs is that they can compute far more query types that human IVR personnel, and, in theory deliver tailored answers for each customer profile.
Beyond the chatbots using popular IM methods such as Facebook Messenger, we foresee that most banks’ AI solutions will extend their functionality to the fast-growing market of home voice-driven personal AI assistants such as Amazon Echo, Google Home and Apple HomePod.
In the US, Capital One has become the first retail bank to allow their customers to track their spending and manage their home and auto loans via simple voice commands using Amazon’s Alexa. No doubt more banks will look to follow their lead and add greater functionality in due course.
Privacy at stake
If banks want to use voice interactive services via home assistant appliances they will likely have to add an additional layer of privacy protection than what is available currently.
In addition to AI, we also see Machine Learning taking a key role in
It is critical for banks to provide the simplest and smoothest user experience while at the same time securing these new channels to protect their sensitive customer information.
From Deskop to Mobile… and now to Home assistant?
Digital channels are constantly evolving and today smartphones are the preferred and dominant platform for financial services. Last year mobile displaced the desktop browser as the main device for accessing financial services.
People expect the mobile experience to be smooth and secure. We’ve seen how banks have embraced sending text messages via IM or SMS as a two-factor authentication technique. Now the next level of interaction is becoming possible as chatbots increase their sophistication. Leaving the handset behind and moving to a fluid conversation with an AI bot is a further improvement in the customer experience.
Today’s investments in AI by banks suggest they are working towards launching services on home assistants. We will be there to ensure user authentication can be performed on such devices either by voice recognition or OOB with the user’s smartphone close to the Home Assistant.
The big question is whether voice control and AI interaction will come to replace our screens as the go-to method of communicating with financial services. We’ll likely still to use screens to get an overview of our accounts, but in terms of payments, transferring money, and learning about new accounts, speaking to an AI is a very attractive proposition.