The Bank of Montreal (BMO) has introduced a new agentic AI tool designed to help contact centre staff resolve credit card enquiries more quickly as part of its efforts to expand the use of AI across customer service.

Its new AI Agent Assist tool supports employees in BMO's Virtual Connect contact centre by bringing together transaction information, internal policies and case guidance in one place. The bank said the technology will reduce the time needed to handle eligible customer enquiries by between 20 and 30 per cent.

BMO receives more than 500,000 calls each year from Canadian customers about credit card transactions, with calls currently taking an average of 12 to 14 minutes to resolve.

The bank said it expects the technology to process around 40 per cent of eligible credit refund requests automatically, allowing some customers to receive refunds on the same day.

The bank said the tool uses a "human-in-the-loop" approach, with employees reviewing AI-generated information before making decisions. Rather than replacing staff, the technology aims to reduce time spent switching between systems and carrying out manual tasks.

BMO said it designed the underlying capabilities so they can support additional customer journeys and internal processes across the bank as it expands its use of agentic AI.

Greg Brown, head of enterprise AI transformation and chief digital officer, personal and business banking Canada at BMO, said: "We're focused on applying AI where it can deliver meaningful, measurable outcomes for our clients and our business.

"We're bringing agentic AI to a high-volume client experience to make it faster and simpler, while building reusable capabilities that support the next generation of AI solutions across BMO. This is how we broaden from individual use cases to AI at scale with clear guardrails and our people accountable for every decision."

The initiative forms part of BMO's wider AI strategy, which aims to generate more than 1 billion Canadian dollars (£530 million) in AI-enabled revenue growth, productivity improvements and operational efficiencies by 2030.

---