Starling Bank has become the first UK bank to integrate a Generative AI chatbot within its digital app, backed by Google’s Gemini Large Language Model (LLM).
Starling’s Spending Intelligence LLM AI chatbot launched today (June 10) to help personal and business account customers gain more sophisticated and efficient spending insights, such as how much a customer has spent this month.
Customers will be able to access Spending Intelligence within a search bar of the ‘spending’ category of the Starling banking app, withoutany additional costs to use.
Starling stated the AI chatbot does not store sensitive information that pertains to the questions asked by customers. Spending Intelligence also does not tell customers what to do with their money; rather gives insights into spending patterns, where money is being spent, etc.
Declan Ferguson, Chief of Finance at Starling, said: “[We’re] Really pleased to be the first UK Bank putting Generative AI in the hands of our customers. This feature, powered by Google’s Gemini, is our first step towards a bigger ambition to implement AI within our banking app to all customers.”
Does Gen AI hold the answer?
Generative AI has exploded in usage since the launch of OpenAI’s ChatGPT, which was seen as similar to LLMs such as Google’s Gemini to provide sophisticated communication to users on any piece of information they may need.
Chatbots have long been in use by traditional banks for the past decade plus, but the onset of GenAI’s capabilities provides greater insights and its digital-native banks like Starling are now taking advantage of the technology.
The integration of Gemini to the Spending Intelligence chatbot is the latest innovation stemming from the Starling and Google partnership, which started in 2018.
Google has provided its Cloud infrastructure to support Starling’s digital banking infrastructure, as well as providing AI tools that help prevent fraud, cyber attacks and assist with customer service.
Now utilising Gemini in a UK-first partnership, Starling is offering deeper insights into customer queries regarding their spending. Customers can customise their questions and filter them specifically to dates and times when a payment was settled.
These types of specific questions were often not capable in the past with past AI chatbots, often being limited to providing background information and links for further customer queries.
