Market · AI in practice
BMO puts an agent on card inquiries and aims to reuse the platform
On November 6, 2026, BMO described AI Agent Assist in Virtual Connect: matching, policy and case prep with a human decision; the time and automation figures are targets, not measured results.
What the sources report
On November 6, 2026, Bank of Montreal said it is using agentic AI in Virtual Connect for credit card transaction inquiries. The centre takes more than 500,000 calls a year on those transactions, averaging 12–14 minutes to resolve. AI Agent Assist handles transaction matching, policy retrieval and case prep; the employee reviews and decides.
Official targets: 20–30% less time on eligible calls and about 40% of eligible credit refund requests processed automatically, with same-day credit possible for qualifying cases. BMO also says it is building these capabilities for reuse across other client experiences and processes.
Adriano's reading
The thesis that matters to me: the value that scales sits in reusable components — matching, policy, case assembly — becoming a platform for the next agents. In high-volume work with clear rules, I would measure success by how often those blocks are reused and by the quality of the human decision at the final gate, alongside time per call. The figures above are published targets, not measured results.