Royal Bank of Canada is deploying AI throughout its operations to drive productivity and to strengthen its position moving forward, both of which are increasing its token usage.  

With increasing use of AI, the bank reported its AI token usage increased by 500% in its fiscal second quarter, ended April 30, Chief Executive Dave McKay said during the bank’s earnings call today. No metric was provided on how many tokens the bank burns on a yearly basis. 

Token usage and cost are rising, Sai Rangachari, chief product officer at core banking provider Temenos, told FinAi News. “As the models become smarter, they tend to think more and burn more tokens.” 

A token is the basic unit of AI model processing; tokens are the small chunks of chatbot inputs and outputs in the form of words and text. That means the cost of an AI workflow scales directly with how much work the AI model performs. 

Burning tokens 

The bank aims to generate nearly CA$1 billion ($720 million) in value from AI by the end of 2027, McKay said last year. 

“We’ve developed over 200 leading edge AI models, rethinking how we operate streamlined workflows and delivering more hyperpersonalized client experiences,” McKay said today. 

AI will help the $1.1 trillion bank serve more clients at no additional cost, he said.  

The technology is “going to make our employees better and it’s going to make our employees more effective in front of the customer,” McKay said. 

For example, the bank’s chatbot “uses AI for intent detection and orchestration, navigating clients to digital capabilities or the best adviser across the network, allowing our people to focus on deepening client relationships,” he said. 

While the tech enables personalized experiences for clients, it also drives internal efficiencies. 

The AI-powered internal chatbot processes about 2 million searches per month concerning bank policies and rules, McKay said. 

“To date, AI has contributed to the development of over 24 million lines of code and facilitated over 120,000 code reviews,” McKay said. 

The macro 

Although the financial services and tech sectors are running full steam on AI development and deployment, McKay noted there are headwinds that can stymie AI and the economy at large. 

“Equity markets are hitting record highs, driven in part by expectations of rising corporate profits and an AI-enabled future,” McKay said.  

“At the same time, bond yields tell a different story, reflecting the risk of monetary tightening as inflation pressures build from both the direct and indirect impact of the energy shock throughout this period of volatility.” 

Register here for the FinAi Lending Summit, set for Oct. 7-8 in Las Vegas.