Quantum Machine Learning Crosses a Banking Milestone as BMO Shows Quantum Utility on IBM Hardware
Machine learning has a new proof point, and it comes from a bank, not a lab. On November 19, 2026, BMO announced it has demonstrated quantum utility in a machine learning application using IBM quantum computers — meaning the work was performed reliably at a scale beyond what exact traditional computing methods can practically reproduce. The project tackles one of machine learning’s most unglamorous but most important problems: deciding which information in a complex dataset actually matters. And it does it with far less data than the old way.
What BMO Actually Demonstrated
The announcement, carried by PR Newswire on November 19, 2026, describes work developed in collaboration with the Democritus University of Thrace (DUTH), run on IBM quantum computers including the latest IBM Quantum Nighthawk r2 processor, and reviewed by IBM’s quantum research team. The application reimagines machine learning feature selection — the step where a model decides which variables in a dataset are worth paying attention to — by adapting the Proxy Quantum Games framework and turning a traditional statistical problem into a dynamic, quantum strategic game.
“Demonstrating quantum utility is an important milestone for BMO and an exciting step forward for the broader financial services industry,” said Kristin Milchanowski, Chief AI and Quantum Officer and Director of the BMO Institute for Applied Artificial Intelligence & Quantum. “While we remain in the early stages of application development, these projects show how quantum computing can help solve business problems in new ways.”
BMO is positioning the achievement as part of a broader strategy to responsibly apply quantum technologies to AI and machine learning, and the bank notes it is among a small group of global financial institutions exploring real-world quantum applications in the near term.
The Quantum Game Behind the Feature Selection
The method itself is the interesting part. In BMO and DUTH’s approach, each feature of a dataset becomes a “player” in a quantum game. A feature scores high when it moves together with the application’s target — the outcome the model is trying to predict — and low when it is uninformative or simply repeats what other features already say. The scores rank the features, and the whole process runs on a quantum computer.
The payoff is efficiency. According to the announcement, the quantum game-based feature selection produces results of similar quality while requiring one to two orders of magnitude fewer samples than traditional feature selection techniques. In plain terms: the same quality of model, built from a fraction of the data. For banks, where the most useful datasets are often the hardest to assemble — fragmented across systems, restricted by privacy rules, or simply expensive to label — needing less data is not a footnote. It is the difference between a model that can be built and one that cannot.
“I’m excited to see BMO’s progress toward real-world quantum use cases in the financial services industry, using our quantum computers to explore novel algorithmic approaches,” said Scott Crowder, Vice President of IBM Quantum Adoption and Business Development.
Why Banks Care: Less Data, Better Models
The potential banking applications BMO names are concrete: more accurate client modeling, improved risk assessment and detection, and more personalized client experiences. Each of those depends on feature selection done well. A risk model that picks the wrong signals misses fraud or misprices credit; a client model built on noisy inputs personalizes badly. If quantum feature selection can find the right signals with far fewer samples, the models downstream get better — and cheaper to train.
The work also fits a longer arc. BMO joined the IBM Quantum Network in February 2025 and continues to invest in quantum research, talent development, and real-world applications, including partnerships with the Chicago Quantum Exchange and Quantum Industry Canada. The bank also launched the BMO Institute for Applied Artificial Intelligence & Quantum, an enterprise-wide Centre of Excellence focused on the responsible application and governance of AI and the advancement of quantum capabilities.
Independent rankings suggest the investment is registering. BMO ranked ninth among financial institutions globally in the 2026 Quantum Index Finance ranking — an independent benchmark of enterprise quantum readiness — and is the highest-ranking Canadian commercial bank in the index, which measures quantum workforce, research and development, and adoption. Separately, Evident AI recognized BMO as one of the world’s top 20 banks in AI and ranked it number two among 50 banks for AI use-case outcomes, the highest-ranking Canadian bank in that category, with the top ranking in the Responsible AI Process indicator group.
A Week of Machine Learning Milestones
BMO’s announcement lands in a week crowded with machine learning news from the big labs. Google DeepMind released Gemini 4 Neon, priced at $3 per million input tokens and $15 per million output tokens, which scored 61 on the Artificial Analysis Intelligence Index — one point above its sibling Carbon — while costing less per completed task, according to reporting in the November 19 AI News Brief. A day earlier, OpenAI released GPT-6.2 Astra and Sol; Astra is the company’s first externally released model to reach its High threshold for AI self-improvement, and both models reached the Critical threshold for cybersecurity capabilities, according to The Neuron’s November 18 briefing. And Anthropic’s IPO, priced at $83 a share, is set to begin trading on the Nasdaq today under the ticker ANT.
The pattern across these stories is consistent: the machine learning field is shifting from raw capability to practical delivery — cheaper completed tasks, models that can accelerate their own research, and, now, quantum techniques that squeeze more signal from less data.
What Comes Next
BMO is careful not to oversell the moment. The bank says it remains in the early stages of application development, and the announcement frames quantum utility as a milestone on the road to practical use in financial services, not a product launch. “At BMO, innovation is about translating emerging technologies into meaningful business impact,” said Steve Tennyson, Chief Technology and Operations Officer at BMO. “This milestone reflects years of investment in innovation, quantum computing, artificial intelligence and strategic partnerships.”
For context on the scale of the institution doing this work: BMO Financial Group is the eighth largest bank in North America by assets, with total assets of $1.5 trillion as of July 31, 2026, and has served clients for more than 200 years.
What to watch next is whether the sample-efficiency advantage holds up as the method moves from demonstration to production workloads — and whether other banks follow with their own quantum machine learning results. If they do, feature selection may become the first place quantum computing earns its keep in finance: not by replacing classical machine learning, but by making it work with far less data.
The Bottom Line
Quantum computing’s promise to machine learning has mostly been theoretical — until a bank showed it working at a scale classical methods cannot practically reproduce. BMO’s quantum feature selection does not claim to beat classical models on quality; it claims to match them while using one to two orders of magnitude fewer samples, and that efficiency is exactly what makes real-world banking models possible. In a week when Google, OpenAI, and Anthropic all made their own machine learning moves, the most practical milestone may have come from a Canadian bank: less data, same quality, real hardware. That is what quantum utility looks like.
Sources: PR Newswire / BMO Financial Group, “BMO Demonstrates Quantum Utility in Machine Learning Application Using IBM Quantum Computers,” November 19, 2026 (via Morningstar); AI News Brief, “AI News Brief — 2026-11-19,” November 19, 2026; The Neuron, “Everything That Happened in AI Today — Wednesday, November 18,” November 18, 2026; AI Briefing, “AI News Today — Daily AI Briefing,” November 19, 2026.
Edited by Muse @QUE.COM
Website: https://QUE.COM Intelligence
Sponsored by: https://MAJ.COM AI Autonomous
Never miss a customer call again — MAJ.COM’s AI receptionist answers every call, texts customers back instantly, and books the appointment for you. See how it works at https://MAJ.COM
Discover more from QUE.com
Subscribe to get the latest posts sent to your email.
