BMO to Deploy Quantum Computing Technology by Year’s End, Says Chief AI Officer

BMO Financial Group, one of the largest banks in North America with $1.4 trillion in total assets, is getting ready to reap the benefits of quantum computing by the end of the year.

That’s according to Kristin Milchanowski, BMO’s chief artificial intelligence (AI) and data officer.

Last month, BMO — the parent of BMO Capital and BMO Wealth Management —  joined the IBM Quantum Network to gain access to its advanced quantum infrastructure.

This would enable BMO to develop and deploy quantum-powered solutions across its operations, such as new approaches to optimize investment portfolio strategies and reveal deep insights into risk management.

“We are in quantum because we do believe that it holds transformational potential,” Milchanowski said in an interview with PYMNTS. “It’s really about preparedness and … getting the muscle memory in place so that we can have first-mover advantage.”

While not ready for prime time, quantum computing nevertheless has hit a turning point. In the past two years, quantum computing has reached a state of utility, she said. That’s the point where it can provide useful results for certain problems, even though it is not yet more broadly useful than today’s standard “classical” computers.

“I can run algorithms and use cases inside a quantum environment and get a useful result — and that’s a significant foot forward,” Milchanowski said.

BMO chose the IBM network because the tech company has “never missed a milestone on their published roadmap, and I really do believe by the end of the calendar year, they will have enough compute power in their quantum machine for me to run really useful algorithms,” she said.

Last year, Fast Company listed BMO as among the world’s most innovative companies in the personal finance category. Also, the company was named the ninth most innovative company in the Innovation in Banking Index by Evident AI. Helping its ranking was placing third in AI-related patents: BMO said its patents were 2.4 times more likely to have forward citations per patent, which is a measure of a patent’s value and impact.

Milchanowski said BMO takes a digital-first strategy in its operations. But when it comes to generative AI, the current form is not scalable for corporations.

“It’s limited and doesn’t scale well in corporate. So that’s why you’re seeing a lot of the hype being busted in the news lately, just because it really doesn’t scale in the version that it’s in today,” Milchanowski said. “But it will add reasoning soon, and that next generation of generative AI will be more useful.”

Generative AI carries high computational costs — even though fees have been coming down — suffers from data and compliance issues, and still hallucinates.

According to OpenAI, its most accurate model GPT-4.5 is only 62.5% factually accurate on average, based on the SimpleQA Accuracy benchmark. GPT-4o, OpenAI o1 and OpenAI o3-mini scored far lower, at 38.2%, 47% and 15% on average.

At present, Milchanowski is focusing on AI agents — bots that do tasks for the user without constant human supervision — in the following areas:

• Chatbot to help customers understand their portfolios more intimately.
• AI agent in the area of pricing and risk management for BMO’s capital markets team. It gathers data from various pricing models, reasons and then a human employee asks the agent to take action. The agent learns from feedback and its own experience.
• Rover AI, a digital assistant, for streamlining the field underwriting process in BMO’s individual life insurance business.
• AI analysis of customer call transcripts to identify pain points and improve communication.
• AI used for automating internal compliance processes, ensuring accuracy and efficiency in customer interactions.

Longstanding AI tools include BMO’s CashTrack Insight, which predicts potential cash shortfalls, such as an account balance dropping below zero, up to seven days in advance. It looks at patterns in spending, regular payments, and income to make these predictions.

Another is the AI-powered BMO Savings Amplifier that lets customers open new savings accounts, set personalized goals, establish automated transfers, and monitor their progress. It uses AI from Personetics to analyze financial data in real time to understand customer behavior, anticipate needs and deliver a personalized experience.

But Milchanowski said these agents have limited autonomy, given the regulatory restrictions in the financial services industry. The agents need human input and they are specifically trained for certain tasks.

“We’re really focused right now on the AI agents, because in banking … everything we do needs to be governed. We do it responsibly, and we want to have a human in the loop,” Milchanowski said. “So these AI agents are really great for us. We’re wanting to introduce them into a more broad customer support function, but we have to be really careful.”

Milchanowski said BMO is working on an AI tool that can deliver a hyper-personalized experience to clients to do things such as adjust financial advice or investment strategies based on real-time market information.

Notably, the bank’s AI efforts are overseen by its governance committee to ensure they are deployed responsibly, Milchanowski said.

When unleashing digital transformation with AI, Milchanowski has this advice to business leaders:

1. Become a steward of understanding innovation, what great companies do to innovate and what are their key tenets. It’s important to adopt an innovative mindset.
2. Eliminate an outdated business process at least once a month, because then it’s easier to implement new technology.
3. Be bold in an era of exponential technological change.

The post BMO to Deploy Quantum Computing Technology by Year’s End, Says Chief AI Officer appeared first on PYMNTS.com.