BMO’s AI agent targets 20–30% faster card support on eligible calls out of 500,000 a year
BMO is bringing an AI agent into its credit-card support centre to help staff answer faster — targeting 20–30% quicker resolutions on eligible inquiries, out of 500,000-plus calls a year. Those figures are targets, not results yet.
📊 Scorecard: our Bullish call on $BMO is scored from $171.13, the last close before it ran. Its first result comes after the next close.
The News
BMO announced on November 6 that it is bringing an AI agent into its credit-card support centre — not to talk to customers, but to help its own staff answer them faster. The tool, called AI Agent Assist, works at BMO’s Virtual Connect contact centre, which handles Canadian clients’ credit-card transaction questions. It brings AI-powered transaction matching, policy lookup and case preparation straight to employees, who review the AI’s work before anything is decided. A human stays in charge: BMO says its client service team remains accountable for reviewing AI-generated information and making servicing decisions.
The scale is what makes the story. Virtual Connect takes more than 500,000 such calls a year — roughly 1,370 a day if spread evenly across the year — and each one currently takes an average of 12 to 14 minutes to resolve. BMO aims to cut the time spent resolving eligible inquiries by 20 to 30%, and to process about 40% of eligible credit refund requests automatically — so qualifying clients could get their credit the same day instead of waiting.
Read the numbers carefully: the 20–30% and 40% figures are targets for eligible calls only, not measured results. No rollout results have been reported yet, and this is employee-assist tooling, not a customer-facing agent. The bigger frame: BMO is designing the underlying capabilities for reuse across the bank, supporting its goal of more than $1 billion in AI-enabled revenue growth, productivity improvements and operating efficiencies by 2030. The 2026 Evident AI Index ranked BMO second among 50 banks for reported AI use-case outcomes.
Targets, not results — but a named bank, a named call centre and 500,000 calls a year is a deployment, not a demo.
Why It Matters
Call centres are where AI agents meet their hardest test: real customers, real money, and a clock. A 12-to-14-minute call is a long time to spend on a credit-card question, and 500,000 of them a year is a lot of staff hours. If BMO’s tool hits even the low end of its target, that is a meaningful saving — and a same-day refund, instead of a wait, is the kind of thing a customer actually notices.
It matters beyond BMO because this is the deployment pattern the whole banking industry is watching: an AI assistant that prepares the work and a human who signs it off. That design sidesteps the trust problem that sank fully automated customer service — the customer still talks to a person, and the person has better information. The reuse point matters too: a tool built once for credit cards can be pointed at mortgages, wealth or business banking without starting over, and that is where the $1 billion goal comes from. The Evident AI Index ranking — second of 50 banks — says BMO is further along than most of its peers at turning AI projects into reported outcomes.
Analyst's Take
Our call is Bullish. The reason is not the percentages — those are targets, and targets are cheap. The reason is the shape of the deployment: a named bank, a named call centre, a concrete volume (500,000-plus calls a year), a human-in-the-loop design, and a plan to reuse the same capabilities across the bank. That is a deployment, not a demo, and the $1 billion-by-2030 goal gives it a number to be judged against.
The honest caveat is the one BMO itself invites: nothing has been measured yet. The test is whether the 20–30% target shows up in real call times, and whether the 40% automatic-refund figure survives contact with real customers. The ranking helps too: second of 50 banks in the Evident AI Index is an outside check, not BMO grading its own homework. Watch for BMO’s first reported results — if the targets hold, every other Canadian bank will be asked why it is not doing the same.
Are you bullish or bearish on this story?
By The Numbers
FAQ
What is AI Agent Assist, in plain terms?
An AI helper for BMO’s own call-centre staff. When a customer calls about a credit-card transaction, the AI finds the matching transaction, pulls up the relevant policy and prepares the case — then a human employee reviews it and makes the decision.
Is the AI talking to customers?
No. This is employee-assist tooling: the customer talks to a person, and the AI helps that person work faster. BMO calls this a human-in-the-loop design.
Are the 20–30% faster resolutions real yet?
No. They are BMO’s targets for eligible inquiries, not measured results. No rollout results have been reported yet.
What does “same-day credit” mean?
If about 40% of eligible refund requests are processed automatically, qualifying clients could see the credit on their account the same day, instead of waiting for manual processing.
What is the Evident AI Index?
An annual ranking of banks on their AI adoption. The 2026 edition ranked BMO second among 50 banks for reported AI use-case outcomes.
👁
Source: BMO / CNW