Bank of Montreal (BMO) is intensifying its artificial intelligence strategy by rolling out a revenue-focused tool deeper into its commercial banking operations. The initiative leverages payments data to provide sales teams with sharper insights into mid-market clients’ financial behaviors, aiming to increase transaction volumes and deepen customer relationships.
The bank has already deployed Codat’s technology in the United States, with a Canadian pilot scheduled in the coming months and a broader rollout targeted for the end of 2026. Codat’s platform connects to more than 20 accounting and enterprise systems, including QuickBooks, Oracle NetSuite, Sage, Microsoft Dynamics, Workday, and Xero. By piping payments and accounts-payable data directly into machine-learning models, the system identifies payment patterns and flags opportunities to shift transactions from slower, higher-risk methods—such as checks—to card-based payments.
Driving Revenue and Efficiency
The logic is straightforward: redirecting check payments to cards enhances client value and boosts bank revenue. According to Joey Rault, Codat’s chief revenue officer, checks still account for 15% to 25% of business payments. Rose Grande, who oversees North American corporate card products at BMO, noted that some clients who adopted the tool increased their BMO payment volumes by approximately 45% after swapping checks for cards. The tool generates tailored recommendations for BMO’s treasury and payments sales teams, enabling more targeted client interactions.
Bradley Leimer, head of Leimer One Advisors, views this as a banking challenge rather than mere tech hype. “Banks want AI to read client behavior and help relationship managers actually deliver,” he said. Commercial banking is a natural fit for AI, Leimer added, because the real value lies in cash flow, payments, working capital, and supplier relationships. However, he cautioned that “the data is a mess.”
Broader AI and Quantum Ambitions
BMO’s AI push extends beyond payments. According to a May 1 report from Bloomberg, the bank is also exploring quantum computing and AI for applications such as earthquake forecasting and wildfire response. Kristin Milchanowski, BMO’s chief AI and quantum officer, has developed a provisional patent for a quantum algorithm designed to predict earthquakes. In April, the bank launched the BMO Institute for Applied Artificial Intelligence & Quantum, with Milchanowski as its first director. Steve Tennyson, head of technology and operations, said the institute aims to channel cutting-edge tech into customer value while managing risk and promoting responsible adoption.
BMO has also joined IBM’s Quantum Network as the first Canadian lender and formed partnerships with Quantum Industry Canada and the Chicago Quantum Exchange, focusing on commercialization, workforce development, and knowledge-sharing.
Market Context and Competition
The timing of BMO’s AI expansion aligns with a more favorable Canadian bond environment. Robert Kavcic, an economist at BMO Capital Markets, noted that long provincial bonds posted gains over the past month, with Government of Canada yields stable and provincial spreads narrowing by about 8 basis points in April. However, investors continue to price in potential Bank of Canada rate hikes due to concerns that high oil prices could feed into core inflation, which might unwind recent support for bond spreads.
Competitors are also advancing their AI capabilities. Wells Fargo’s Ather Williams III said the firm’s AI platform now covers 85% of use cases, with 26 applications already in production after 335 tests. He expects some new payment features to become “table stakes” within five years.
While AI-driven sales leads promise efficiency gains, challenges remain. The tools depend on authorized data access, client trust, and transparency—especially when the software suggests rather than confirms. BMO’s ultimate test will be whether its bankers can convert raw payments data into actionable advice, boost card and treasury activity, and deepen client relationships without customers feeling like mere revenue targets.
