Skip to content
    The Canadian Bank AI Index · 2026 Q4 edition

    Which Canadian banks use AI, and how openly?

    Short answer

    The Canadian Bank AI Index scores 35 Canadian banks, credit unions and fintechs out of 100 on public evidence of AI: what reaches customers, governance, model-risk readiness, transparency and talent. TD leads the 2026 Q4 edition with 84. Every point links to a dated source, and the 210-entry database is free to download.

    Figures as of November 1, 2026.

    Mitey, data editor · Evidence checked Independent. Not affiliated with any bank. No paid placement.

    Scorecard

    The top 12, out of 100

    Five pillars, 20 points each. Bars show where each institution earns its points. Tap a name for the evidence behind every point.

    • Customer deployment
    • Governance disclosure
    • Model-risk readiness
    • Transparency to customers
    • AI talent & capacity
    Canadian Bank AI Index 2026 Q4: top 12 scores out of 100
    1. 84
    2. 81
    3. 80
    4. 69
    5. 57
    6. 55
    7. 54
    8. 42
    9. 37
    10. 30
    11. 24
    12. 24

    All 35 institutions, sortable, with every criterion

    How points are earned

    • Deployment: live customer uses across six areas
    • Governance: principles, board oversight, risk disclosure, an accountable executive
    • Model risk: model-risk controls that cover AI, aligned to OSFI E-23
    • Transparency: customers told; voluntary code; regular reporting
    • Talent: AI lab, disclosed headcount, active hiring
    Full method
    By the numbers

    What the public record shows

    210

    AI deployments in our database

    27 institutions, 2017–2026, every row sourced

    21%

    are used by customers directly

    Another 13% run on customer accounts unseen, like fraud screening

    6 of 35

    publish responsible-AI principles

    13 tell customers when AI or automated decisions are used

    181

    days until OSFI E-23 applies to AI models

    Model-risk rules for federally regulated banks take effect May 1, 2027

    Deployment database

    AI announcements and disclosures by year

    By the year each use launched, was announced or was first disclosed, never by a web page’s last-updated date, and coloured by who the AI actually reaches. Most of what banks announce is for staff or research, not for customers. 5 entries with no published date are not charted.

    Number of sourced AI deployments by year, split by reach2017: 3 Customers use it directly2017: 3 Research & governance2018: 1 Customers use it directly2018: 3 Research & governance2019: 5 Customers use it directly2019: 1 Runs on customer accounts2019: 1 Staff tools2019: 3 Research & governance2020: 5 Customers use it directly2020: 2 Staff tools2020: 1 Research & governance2021: 5 Customers use it directly2021: 3 Runs on customer accounts2021: 3 Staff tools2021: 2 Research & governance2022: 2 Customers use it directly2022: 3 Research & governance2023: 3 Runs on customer accounts2023: 1 Staff tools2023: 2 Research & governance2024: 3 Customers use it directly2024: 2 Runs on customer accounts2024: 14 Staff tools2024: 7 Research & governance2025: 11 Customers use it directly2025: 11 Runs on customer accounts2025: 23 Staff tools2025: 20 Research & governance2026: 7 Customers use it directly2026: 8 Runs on customer accounts2026: 29 Staff tools2026: 18 Research & governance
    Number of sourced AI deployments by year, split by reach
    YearCustomers use it directlyRuns on customer accountsStaff toolsResearch & governance
    20173003
    20181003
    20195113
    20205021
    20215332
    20222003
    20230312
    202432147
    202511112320
    2026782918
    • Customers use it directly
    • Runs on customer accounts
    • Staff tools
    • Research & governance

    Share of all 210 deployments

    Share of deployments by reach
    • Customers use it directly45
    • Runs on customer accounts28
    • Staff tools75
    • Research & governance62
    Search the database
    How the Index is built

    From a bank’s own words to a score you can check

    1. Collect what institutions publish

      Annual reports, filings, newsrooms, product pages, patents, job boards and reputable news: 210 AI uses at 27 institutions, each dated by when it launched and linked to its source.

      Mitey the yeti checks a speech bubble through a magnifying glass
    2. Score five pillars, 20 points each

      Customer deployment, governance, model-risk readiness, transparency and talent. Every point is a yes-or-no criterion met only with a public, dated source, and the same rules apply to every institution.

    3. Show customers what it means

      Pick your bank to see the AI it has disclosed on accounts like yours, your rights, and how to ask a person to review an automated decision.

      Mitey the yeti checks a phone showing a shield with a tick
    Latest launches

    Newest AI that reaches customers

    LaunchedQuestrade

    Flows: plain-language trading automation engine

    Questrade announced Flows, which lets clients describe an investment strategy in plain language and then builds and runs the automation, including placing trades for them; Questrade presents it as part of its 'agentic finance' launch.

    Context: The release does not name the AI model behind Flows; Fintech.ca (2026-07-29) describes it under 'AI-Powered Trading Automations'. Not marked 'coming soon' in the release, but rollout status not verified. 'Canada's first' is Questrade's claim.

    Customers use it directlyWealth & investingLaunchedPress release: Questrade (GlobeNewswire via BNN Bloomberg)
    AnnouncedQuestrade

    Questrade MCP: client accounts connected to Claude and ChatGPT

    Questrade announced an MCP integration that lets clients connect accounts to AI assistants such as Claude and ChatGPT to view and analyse holdings, research markets and initiate orders, which clients confirm in the Questrade app; Questrade calls itself the first Canadian institution to do this officially.

    Context: Beta: NCFA Canada (https://ncfacanada.org/questrade-mcp-ai-brokerage-accounts/) reports Questrade published beta documentation on 2026-07-23 supporting Claude and Claude Code. 'First' is Questrade's claim. The AI models are third-party (the client's own assistant). In-app order confirmation per BetaKit 2026-07-30.

    Customers use it directlyAssistants & chatPilotPress release: Questrade (GlobeNewswire via BNN Bloomberg)
    LaunchedBMO

    SmartDecision AI-enhanced life insurance underwriting

    BMO Insurance launched SmartDecision, which uses predictive modelling and AI-enhanced capabilities to give real-time underwriting decisions on eligible life applications up to $5 million, often without medical requirements.

    Context: Advisor-led application; eligible applications only.

    Runs on customer accountsStaff & operationsLaunchedNewsroom / blog: BMO
    AnnouncedTru (First West)

    DataVisor AI fraud and risk platform via VeriPark

    Tru Cooperative Bank (formerly First West) selected DataVisor's AI-powered fraud and AML platform, delivered through VeriPark's digital banking platform, for real-time fraud detection across onboarding, login, profile changes, Interac e-Transfer and bill pay.

    Context: vendor-issued (DataVisor release via Business Wire, 2026-06-16, not opened). Selection announced; go-live not confirmed.

    Runs on customer accountsFraud & identityAnnouncedNews report: Finovate
    AnnouncedInterac

    Interac Verified: passive liveness and AI-generated media detection (Incode)

    Interac announced it is adding Incode's passive liveness detection and Deepsight, which tests for AI-generated images or video, virtual cameras and device tampering, to its Interac Verified identity service, with a phased rollout targeting Q3 2026.

    Context: Published 2026-05-25, updated 2026-06-15. Whether the Q3 2026 launch happened was not verified. End users of businesses using Interac Verified go through the check.

    Customers use it directlyFraud & identityAnnouncedPress release: Interac Corp.
    LaunchedTD

    Agentic AI for mortgage and HELOC pre-adjudication

    TD launched its first agentic AI system, built by Layer 6, which classifies documents, calculates income and checks applications against policy before producing a memo for human underwriters on mortgage and HELOC applications.

    Context: Underwriters make the decisions. A July 2026 TD Stories article says 100% of new personal mortgages go through the process (TD-reported).

    Runs on customer accountsCredit & lendingLaunchedNewsroom / blog: TD Bank Group

    All 210 entries

    Regulation tracker

    The rules, and when they bite

    ComingFederal (Canada)

    OSFI Guideline E-23 (model risk)

    Final, published September 11, 2025. Takes effect May 1, 2027, and expressly covers AI and machine-learning models.

    E-23 effective for federally regulated financial institutions, May 1, 2027: 181 days to go

    ComingQuebec

    AMF guideline on AI (Quebec)

    Final. The guideline is dated March 2026, the AMF’s guideline page was updated April 24, 2026, and it takes effect May 1, 2027 for institutions the AMF regulates, including Desjardins.

    AMF AI guideline effective, May 1, 2027: 181 days to go

    ComingFederal (Canada)

    Consumer-Driven Banking Act (open banking)

    Royal Assent March 26, 2026, but most provisions are not in force. Regulations were proposed June 27, 2026; final rules are targeted for late 2026 or early 2027. No date yet for data sharing to start.

    ProposedFederal (Canada)

    Bank anti-fraud regulations (proposed)

    Proposed in Canada Gazette Part I on June 27, 2026, with a proposed start date of July 1, 2027. Final version not yet published.

    Proposed coming-into-force date, July 1, 2027: 242 days to go

    All 19 items with dates

    Global context

    How Canada’s biggest banks rank worldwide

    The Evident AI Index ranks the world’s largest banks on AI talent, innovation, leadership and transparency. Only the largest Canadian banks are in it.

    Bank2026of 502025of 502024of 50
    RBC#3#3#3
    TD#10#13#9
    BMO#20#19#24
    CIBC#21#22#22
    Scotiabank#30#29#20

    Global rank among the large banks Evident assesses. Source: Evident AI Index 2026 (Evident (primary); corroborated by The Logic 2026-10-06 and bank press releases); Evident AI Index 2025 (news (The Logic; RBC rank corroborated by Fintech.ca 2025-10-09); Evident's own page for this edition not found); Evident AI Index 2024 (news (The Logic; corroborated by Fintech.ca 2024-10-18); Evident's own page for this edition not found).

    For customers

    What AI touches your account, and your rights

    Pick your bank to see the AI it has disclosed on accounts like yours: fraud screening, chat, credit decisions. Then see what you can ask for, including a human review of an automated decision and how to complain.

    Open data

    Use the data

    Every score, criterion, deployment and regulation item is free to reuse under CC BY 4.0. Credit “Canadian Bank AI Index, AIBanks.ca”, or embed the live ranking on your own site.

    About the Index

    What is the Canadian Bank AI Index?

    An independent scorecard of 35 Canadian banks, credit unions and fintechs, built only from public evidence: where AI reaches customers, how it is governed, whether customers are told, and the talent behind it. Every point links to its source. Read the method.

    Which Canadian bank is furthest ahead on AI?

    On public evidence, TD Bank Group (The Toronto-Dominion Bank) scores highest in this edition, with 84 out of 100. See the scorecard for everyone else. Our score measures what institutions disclose and launch, not whether their AI is good or safe. Globally, the Evident AI Index ranks the large banks on talent, innovation, leadership and transparency.

    Does my bank use AI on my account?

    Almost certainly for fraud screening, and often for chat, credit and marketing. Our “What AI touches my account?” page lists what each institution has disclosed and explains your rights, including how to ask for a human to review a decision.

    Is there an AI bank in Canada?

    Not in the sense of a bank run by AI. Every Canadian bank is run by people and regulated as usual; what differs is how much AI each uses and discloses. Digital banks and fintechs often use AI heavily in onboarding and fraud checks but publish little about it. Compare them on the scorecard.

    How often is the Index updated?

    Quarterly, with new deployments added as they are announced. This is the 2026 Q4 edition, with evidence checked to 2026-11-01. The change log lists every update.

    Get the quarterly Canadian Bank AI Index update

    New scores, new deployments and rule changes, once a quarter from the MITEY Network. Unsubscribe any time.

    Educational information only, not financial, legal or privacy advice. Index scores measure public evidence, not service, safety or fairness. If you think you have been scammed, contact your financial institution right away and report it to the Canadian Anti-Fraud Centre at 1-888-495-8501. Last reviewed November 1, 2026.