Skip to content

    Regulation tracker

    The rules for AI in Canadian banking, and when they bite

    Short answer

    Canada has no law that regulates AI in banking as such. The binding rule coming is OSFI Guideline E-23 on model risk, effective May 1, 2027, alongside Quebec’s AMF AI guideline. Bill C-36, which would add a right to explanations of automated decisions, is not law. This tracker lists 19 items with verified status.

    Figures as of November 1, 2026.

    19 guidelines, laws, bills and codes, each with its current status checked against the official source and a countdown to the next milestone.

    Mitey, data editor · Evidence checked Statuses change. Each item shows the date we last checked it.

    What happens next

    1. May 1, 2027 · 181 days to go

      OSFI Guideline E-23 (model risk): Guideline takes effect

    2. May 1, 2027 · 181 days to go

      AMF guideline on AI (Quebec): Guideline takes effect

    3. July 1, 2027 · 242 days to go

      Bank anti-fraud regulations (proposed): Proposed coming-into-force date

    4. May 15, 2029 · 926 days to go

      Bank anti-fraud regulations (proposed): First annual fraud-data report to FCAC due (covering 2028)

    Recently

    Coming into force

    Final rules with a start date.

    ComingOffice of the Superintendent of Financial Institutions (OSFI) · Federal (Canada)

    OSFI Guideline E-23 – Model Risk Management (2027)

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

    For customers

    The models your bank uses to approve credit, set prices or flag fraud, including AI and machine-learning models, will need documented oversight, bias checks and ongoing monitoring. You will not see the guideline directly, but it is the main federal rule that makes a bank answer for how its algorithms behave.

    For banks

    Federally regulated institutions need an enterprise model inventory, risk ratings and lifecycle controls (design, independent review, approval, deployment, monitoring, decommissioning) that explicitly cover AI/ML, explainability and model drift. Requirements scale with the institution's size and complexity, and heavy use of advanced AI is expected to be matched by more mature governance.

    E-23 effective for federally regulated financial institutions

    May 1, 2027

    181 days to go
    • Sep 11, 2025 Final guideline published
    • May 1, 2027 Guideline takes effect

    Checked November 1, 2026

    ComingAutorité des marchés financiers (AMF) · Quebec

    AMF (Quebec) Guideline on the Use of Artificial Intelligence (Ligne directrice sur l'utilisation de l'intelligence artificielle)

    Status: 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.

    For customers

    The AMF says the guideline is meant to protect consumers, and one of its five sets of expectations is the fair treatment of clients. It applies to any use of AI by a Quebec-regulated institution, whether or not the AI handles client records. It does not apply to federally chartered banks, which follow OSFI.

    For banks

    Quebec-regulated institutions must manage AI systems with a risk-based classification, set institution-wide governance and risk management for AI, document controls for each stage of an AI system’s lifecycle, and meet fair-treatment expectations, all on top of the AMF’s Model Risk Management Guideline, applied in proportion to their size and risk.

    AMF AI guideline effective

    May 1, 2027

    181 days to go
    • Mar 2026 Final Guideline for the Use of Artificial Intelligence (dated March 2026)
    • Apr 24, 2026 AMF guideline page updated
    • May 1, 2027 Guideline takes effect

    Checked November 3, 2026

    ComingDepartment of Finance Canada; Bank of Canada (supervisor) · Federal (Canada)

    Consumer-Driven Banking Act (open banking) and proposed Consumer-Driven Banking Regulations

    Status: 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.

    For customers

    You will eventually be able to share your banking data with accredited apps through secure connections instead of handing over your password for screen scraping, with the Bank of Canada supervising and a public registry of approved providers. Data sharing has not started, and the screen-scraping ban only applies once the framework is fully operational.

    For banks

    Large banks will be participating entities subject to common rules on consent, liability, authentication and recordkeeping, with a staggered rollout that may stretch 18 months or more by product. AI-driven services such as cash-flow underwriting will depend on this data pipeline; the draft regulations do not mention AI.

    Final regulations (targeted late 2026 or early 2027) and Bank of Canada draft guidance (expected fall 2026); accreditation first, data-sharing phased starting with deposit and payment accounts

    • Nov 6, 2025 Budget 2025: Bank of Canada to oversee; write access to be legislated by mid-2027, depending on Real-Time Rail
    • Mar 26, 2026 Re-enacted Act receives Royal Assent (Budget 2025 Implementation Act, No. 1)
    • Jun 27, 2026 Proposed regulations pre-published (60-day comment period)
    • Jul 21, 2026 Bank of Canada advisory committee meeting: draft BoC guidance expected in the fall
    • Aug 26, 2026 Comment period closes

    Checked November 1, 2026

    In force now

    Rules that already apply.

    In forceNational Assembly of Quebec; overseen by the Commission d'accès à l'information (CAI) · Quebec

    Quebec Act respecting the protection of personal information in the private sector, s. 12.1 (automated decision notice, from Law 25)

    Status: In force since September 22, 2023. Whether it binds federally chartered banks is not settled; it applies to Quebec-regulated institutions such as Desjardins.

    For customers

    If a business decides something about you based only on automated processing of your personal information, it must tell you no later than when it tells you the decision. On request it must say what information it used, the reasons and main factors behind the decision, and that you can have the data corrected, and you can present your views to a staff member who can review the decision.

    For banks

    Fully automated decisions such as instant credit declines for Quebec customers need a notice and an explanation-and-review process. Whether s. 12.1 binds federally chartered banks (versus PIPEDA) is a constitutional question not verified here; it clearly applies to Quebec-regulated institutions such as financial cooperatives.

    None — in force

    • Sep 22, 2023 s. 12.1 comes into force

    Checked November 1, 2026

    In forceOSFI · Federal (Canada)

    OSFI Guidelines B-10 (Third-Party Risk Management) and B-13 (Technology and Cyber Risk Management)

    Status: In force. Neither guideline names AI, but OSFI’s July 2026 AI bulletin applies both to generative and agentic AI.

    For customers

    When your bank relies on an outside AI or cloud provider, it is still responsible for keeping your data and service secure and available. These rules are why banks must plan for outages and vendor failures rather than blaming the vendor.

    For banks

    AI vendors and cloud platforms fall under third-party risk management (B-10), including concentration risk, while AI-enabled systems and AI-specific attacks such as prompt injection fall under technology and cyber risk management (B-13). OSFI now expects third parties to disclose their AI use and AI dependencies to be mapped to critical operations.

    No announced revision; watch for AI-specific supervisory expectations building on the July 2026 bulletin

    • Jul 31, 2022 B-13 published
    • Apr 30, 2023 B-10 published
    • Feb 20, 2024 Both pages last modified
    • Jul 13, 2026 OSFI AI bulletin maps gen/agentic AI risks onto B-13 and B-10

    Checked November 1, 2026

    In forceCanadian Securities Administrators (CSA) · Canada (provincial/territorial securities regulators)

    CSA Staff Notice and Consultation 11-348, Applicability of Canadian Securities Laws and the use of Artificial Intelligence Systems in Capital Markets

    Status: Published and effective December 5, 2024. Consultation closed March 31, 2025; no follow-up rule yet.

    For customers

    If a bank-owned dealer or adviser uses AI to recommend investments, the same suitability, conflict-of-interest and fairness duties apply as if a person did it. The CSA discourages 'black box' AI that cannot be explained.

    For banks

    Bank-owned registrants, and the banks as reporting issuers, must apply existing securities law to AI: governance, explainability, and fair, substantiated AI claims in disclosure (no 'AI-washing'). The notice creates no new rules.

    No CSA follow-up rule or summary of comments found

    • Dec 5, 2024 Notice published
    • Mar 31, 2025 Comment deadline

    Checked November 1, 2026

    Proposed

    Bills, drafts and consultations that could change.

    ProposedDepartment of Finance Canada; supervised by FCAC · Federal (Canada)

    Proposed Regulations Amending the Financial Consumer Protection Framework Regulations (bank anti-fraud rules under Bill C-15 Bank Act amendments)

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

    For customers

    Your bank would need your express consent before turning on e-Transfers, wires and similar capabilities, and you could switch some of them off or adjust limits. The rules respond partly to AI-enabled scams such as deepfake videos and AI-generated calls.

    For banks

    Banks would need documented fraud detection and investigation criteria reviewed annually, consumer notification practices, and annual fraud-data reporting to FCAC. The text does not require AI, but fraud-detection systems (often ML-based) become a regulated control.

    Proposed coming-into-force date

    July 1, 2027

    242 days to go
    • Jun 27, 2026 Proposed regulations pre-published
    • Jul 1, 2027 Proposed coming-into-force date
    • May 15, 2029 First annual fraud-data report to FCAC due (covering 2028)

    Checked November 1, 2026

    ProposedParliament of Canada (sponsor: Minister of Artificial Intelligence and Digital Innovation) · Federal (Canada)

    Bill C-36 (45-1), An Act to enact the Protecting Privacy and Consumer Data Act (PPCDA)

    Status: Introduced June 15, 2026. Still at second reading in the House of Commons; not law. It would add a federal right to an explanation of significant automated decisions.

    For customers

    If passed, you could ask a business, including a bank, to explain a prediction, recommendation or decision made about you by an automated system when it could have a legal or similarly significant effect, including what data was used and the main reasons. You would also get a chance to make written representations to an employee who can review the decision, plus a right to erasure.

    For banks

    Banks would need plain-language public accounts of automated decision systems with significant effects, on-request explanations within 30 days, and human review channels, enforced by a new Digital Safety and Data Protection Commission. Penalties could reach the greater of $10M or 3% of global revenue (AMPs) and $25M or 5% (fines).

    Second reading debate and referral to committee (not scheduled)

    • Jun 15, 2026 First reading in the House of Commons
    • Jul 6, 2026 Library of Parliament preliminary legislative summary

    Checked November 1, 2026

    ProposedPrime Minister's Office; Innovation, Science and Economic Development Canada · Federal (Canada)

    Minister of AI and Digital Innovation, National AI Strategy 'AI for All' and federal AI transparency consultation

    Status: Strategy announced June 4, 2026. An AI transparency consultation closed September 23, 2026. No AI bill has been tabled.

    For customers

    The consultation asked whether people should be told when they are interacting with an AI system and how AI-generated content should be identified, which could later shape disclosure rules for bank chatbots. Nothing binding has come from it yet.

    For banks

    Pillar 1 of the strategy promises modernized privacy and online-safety laws (Bills C-36 and C-34) rather than a standalone AI act. Banks should expect transparency, incident-tracking and AI-agent monitoring themes from the consultation to reappear in future policy.

    Government response to AI transparency consultation (no date)

    • May 13, 2025 Evan Solomon named Minister of Artificial Intelligence and Digital Innovation
    • Jun 4, 2026 National AI Strategy 'AI for All' announced (3 principles, 6 pillars)
    • Jul 23, 2026 AI transparency consultation opens
    • Sep 23, 2026 AI transparency consultation closes

    Checked November 1, 2026

    ProposedParliament of Canada (sponsor: Minister of Canadian Identity and Culture) · Federal (Canada)

    Bill C-34 (45-1), Safe Social Media Act (Digital Safety Act and Digital Safety Commission of Canada Act)

    Status: At committee in the House of Commons after second reading on October 29, 2026. Not a banking law, but Bill C-36 relies on the commission it would create.

    For customers

    This bill is about online safety, not banking, but it would create the commission that Bill C-36 relies on to enforce new privacy and automated-decision rights. Its progress therefore affects when any new AI-explanation rights could be enforced.

    For banks

    Direct obligations on banks are not expected from C-34 itself (not verified against bill text). Its importance is indirect: C-36 renames and uses the C-34 Commission as the privacy regulator.

    Committee study (no meetings recorded yet)

    • Jun 10, 2026 First reading
    • Oct 29, 2026 Second reading; referred to Canadian Heritage committee

    Checked November 1, 2026

    Voluntary

    Codes institutions can sign up to.

    VoluntaryInnovation, Science and Economic Development Canada (ISED) · Federal (Canada)

    Voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI Systems

    Status: Open for signature since September 2023. CIBC signed on March 25, 2025, the first Big Six bank to do so; Interac signed in November 2024.

    For customers

    A signatory bank has publicly promised to follow principles such as transparency, human oversight and safety when building or managing advanced generative AI. The code is voluntary and has no enforcement, so it is a signal, not a right you can claim.

    For banks

    Signing is a low-cost public trust signal; only CIBC among the large banks has done it, so peers' absence is visible. FCAC's consumer AI page cites the code as one of the government's voluntary measures.

    No scheduled milestone; watch for other banks signing

    • Dec 7, 2023 8 more signatories incl. CGI, IBM (no banks)
    • May 27, 2024 8 more signatories incl. Mastercard (no banks)
    • Nov 8, 2024 10 more incl. Interac Corp. and Nuvei; total 40 signatories
    • Mar 25, 2025 CIBC signs; says it is the first major Canadian bank to do so

    Checked November 1, 2026

    VoluntaryOffice of the Privacy Commissioner of Canada with provincial and territorial privacy commissioners · Federal, provincial and territorial (Canada)

    Principles for responsible, trustworthy and privacy-protective generative AI technologies

    Status: Issued December 7, 2023 by federal, provincial and territorial privacy regulators. Non-binding guidance on how existing privacy law applies.

    For customers

    Privacy regulators expect you to be told when you are talking to a generative AI tool such as a chatbot and when generative AI is used in a decision about you, and to be able to challenge significant decisions and ask for human review. Access to finance is named as a high-impact context needing extra care.

    For banks

    Banks deploying generative AI should be able to show legal authority and consent, necessity and proportionality, openness, accountability and accuracy checks under nine headings. Accountability for decisions stays with the organization, not the AI tool.

    None scheduled

    • Dec 7, 2023 Joint principles issued

    Checked November 1, 2026

    Stalled or died

    Efforts that did not become law, kept for context.

    Stalled / diedParliament of Canada (Minister of Innovation, Science and Industry) · Federal (Canada)

    Bill C-27 (44-1), Digital Charter Implementation Act, 2022 (CPPA, Tribunal Act, Artificial Intelligence and Data Act)

    Status: Died on the Order Paper when Parliament was prorogued on January 6, 2025. The Artificial Intelligence and Data Act has not been reintroduced.

    For customers

    Canada never passed AIDA, so there is still no federal law that regulates AI systems as such. Your protections come from privacy law, banking consumer-protection rules and regulator guidance instead.

    For banks

    There is no federal AI statute to comply with; AI obligations come through OSFI guidance, privacy law and consumer-protection rules. AIDA-style obligations could return in a future bill, but none is before Parliament.

    None — AIDA has not been reintroduced; privacy parts largely carried into Bill C-36 (45-1)

    • Jun 16, 2022 Introduced (first reading)
    • Apr 24, 2023 Second reading and referral to committee
    • Jan 6, 2025 Session ends (prorogation); bill dies

    Checked November 1, 2026

    Regulator reports

    Not rules, but signals of what supervisors expect.

    ReportOSFI · Federal (Canada)

    OSFI Technology Risk Bulletin on Generative and Agentic AI, fall 2026 Risk Outlook on frontier AI, and statement endorsing FSB Sound Practices for AI

    Status: July 2026 bulletin on generative and agentic AI; October 8, 2026 risk outlook focused on frontier AI; October 15, 2026 endorsement of the FSB’s AI sound practices. Non-binding.

    For customers

    OSFI is telling banks to keep a human accountable for important AI-assisted decisions and to tightly limit what AI agents can do inside bank systems. It does not give you new rights, but it signals that the regulator expects human oversight of AI that affects you.

    For banks

    Expectations include senior-management AI literacy, integrating AI risk into enterprise risk management, unique least-privilege identities and logging for AI agents, validation of AI-generated code, prompt-injection threat modelling and manual fallbacks for AI outages. These are framed as sound practices complementing B-10, B-13, E-21 and E-23 rather than new binding rules, but boards and senior management remain accountable.

    OSFI Annual Risk Outlook (published each spring; 2027 date not announced)

    • Jul 13, 2026 Generative and Agentic AI technology risk bulletin (page modified date; bulletin dated July 2026)
    • Oct 8, 2026 Fall 2026 Semi-Annual Risk Outlook focused on frontier AI
    • Oct 15, 2026 OSFI endorses FSB Sound Practices for Responsible Adoption of AI (12 practices)

    Checked November 1, 2026

    ReportOSFI and Financial Consumer Agency of Canada (FCAC) · Federal (Canada)

    OSFI-FCAC Risk Report – AI Uses and Risks at Federally Regulated Financial Institutions

    Status: Published September 24, 2024. Not formal guidance, but it sets out the risks supervisors expect banks to manage.

    For customers

    Regulators have told banks to watch AI for bias in credit and pricing decisions, keep humans in the loop and be open with customers about how AI and their data are used. It also warns that AI makes scams such as deepfakes and voice spoofing cheaper to run.

    For banks

    The report sets the supervisory baseline: about 50% of institutions used AI in 2023 with 70% expected by 2026, and risks span data governance, explainability, third-party concentration, cyber and AI-enabled fraud. Recommendations include treating AI as a cross-cutting risk, engaging boards and having an AI strategy even if not adopting AI.

    No follow-up edition announced

    • Sep 24, 2024 Report published

    Checked November 1, 2026

    ReportOSFI and Global Risk Institute; FIFAI II also sponsored by Finance Canada, FINTRAC, FCAC and Bank of Canada · Federal (Canada)

    Financial Industry Forum on Artificial Intelligence (FIFAI I 'EDGE' report and FIFAI II 'AGILE' final report)

    Status: FIFAI II final report published March 23, 2026, proposing the AGILE framework. Explicitly not regulatory guidance.

    For customers

    The forum recommends that banks tell you when you are dealing with AI, explain material AI-driven decisions where feasible, offer opt-outs where possible and give you a timely way to get redress. These are recommendations, not legal rights.

    For banks

    FIFAI II is the clearest statement of what Canadian regulators and industry consider responsible AI practice, including human oversight of material AI-assisted decisions and bias testing for consumer-facing AI at critical third parties. It carries no legal force but is a likely reference point for future supervisory expectations.

    No FIFAI III announced

    • Apr 30, 2023 FIFAI I report 'A Canadian Perspective on Responsible AI' (EDGE: Explainability, Data, Governance, Ethics)
    • Mar 23, 2026 FIFAI II final report (AGILE: Awareness, Guardrails, Innovation, Learning, Ecosystem resiliency)
    • Mar 25, 2026 FCAC announces FIFAI II final report

    Checked November 1, 2026

    ReportFinancial Consumer Agency of Canada (FCAC) · Federal (Canada)

    FCAC consumer guidance 'Artificial intelligence in banking' and FCAC Strategic Plan 2026-2029

    Status: Consumer page on AI in banking dated October 27, 2026. FCAC has no AI-specific rule for banks or chatbots.

    For customers

    FCAC says Canada has no laws that apply only to AI, but existing rules still apply when banks use chatbots or AI: products must suit your needs, information must be clear and not misleading, and privacy law applies. Complaints about AI-driven service go through the normal route: your bank first, then the external complaints body (OBSI).

    For banks

    AI-delivered advice, sales and service are supervised under existing market-conduct rules such as the appropriate-products guideline and complaint-handling requirements. FCAC's plan signals supervisory interest in opaque or unfair automated decision-making.

    No AI-specific FCAC guidance announced

    • Jul 21, 2026 Strategic Plan 2026-2029 page date (document dated March 2026)
    • Oct 27, 2026 'Artificial intelligence in banking' consumer page date

    Checked November 1, 2026

    ReportCompetition Bureau Canada · Federal (Canada)

    Competition Bureau work on AI and algorithmic pricing

    Status: ‘What We Heard’ report on algorithmic pricing published January 22, 2026. No enforcement or rule change announced.

    For customers

    The Bureau is watching whether shared pricing algorithms could quietly push prices up or reduce choice. Banking is not named, but the concern applies to any sector that sets rates or fees with algorithms.

    For banks

    Using common third-party pricing models or data could raise Competition Act questions if it facilitates coordination. No bank-specific guidance exists.

    No next step scheduled; Bureau says it will keep engaging

    • Jun 10, 2025 Algorithmic pricing consultation opens
    • Aug 4, 2025 Consultation closes (100+ submissions)
    • Jan 22, 2026 'What We Heard' report on algorithmic pricing

    Checked November 1, 2026

    ReportFinancial Transactions and Reports Analysis Centre of Canada (FINTRAC) · Federal (Canada)

    FINTRAC and AI (anti-money laundering)

    Status: No AI-specific obligations for banks. FINTRAC co-sponsored the FIFAI II report and describes its own AI use.

    For customers

    Banks' AML monitoring of your transactions increasingly uses machine learning, and FINTRAC itself is adopting AI to analyse reports. There are no new customer-facing rights here.

    For banks

    Existing PCMLTFA reporting, record-keeping and know-your-client duties apply regardless of whether AI is used. FINTRAC plans conversational and agentic AI and data/AI governance frameworks internally, which may change how it engages reporting entities.

    None announced

    • Mar 9, 2026 2026-27 Departmental Plan (modified date)
    • Mar 23, 2026 FINTRAC co-sponsors FIFAI II final report

    Checked November 1, 2026

    Regulation questions

    Is there a Canadian law that regulates AI at banks?

    Not a general AI law. The federal Artificial Intelligence and Data Act died when Parliament was prorogued in January 2025. Banks are governed by existing law (privacy, consumer protection, human rights) and by OSFI guidance, most importantly Guideline E-23 on model risk, which covers AI models from May 1, 2027.

    What changes for customers when E-23 takes effect?

    Nothing you will see directly. E-23 requires federally regulated institutions to inventory, rate, validate and monitor their models, including AI. The aim is fewer harmful errors, such as biased or drifting credit and fraud models.

    Do I have a right to know when AI decides about me?

    In Quebec, yes: since September 22, 2023, an organization that makes a decision based exclusively on automated processing must tell you and, on request, explain it and let you have it reviewed by a person. Elsewhere, you can ask for the reasons and the information used through your privacy access rights. See What AI touches my account?.

    Educational information only, not financial, legal or privacy advice. Rules and bank features change; check the linked official pages and talk to your financial institution or a qualified professional about your situation.

    How many of these rules bind banks today

    Each item is checked against its official source and placed in one stage.

    How many of these rules bind banks today: inputs, steps and result
    Items tracked19
    In force3
    Final but not yet in effect3
    Proposed or stalled5
    Voluntary codes and reports8
    Binding today3 of 193 more take effect later

    Open the next rule to take effect →

    Mitey the yeti at a desk with a laptop and a dated checklist
    Mitey keeps the countdown to May 1, 2027, when OSFI E-23 and the AMF AI guideline both take effect.

    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.