Skip to content

    Caisse network · Regulated by provincial (Autorité des marchés financiers, Québec)

    Desjardins Group (Mouvement Desjardins): AI score and deployments

    Short answer

    Yes. Desjardins has 8 sourced AI entries in our database, 1 of them reaching customers, and scores 54 out of 100 (Developing) in the Canadian Bank AI Index, rank 7 of 35. Its strongest pillar is model-risk readiness (20/20); its weakest is governance disclosure (5/20).

    Figures as of November 1, 2026.

    North America’s largest financial cooperative, regulated mainly by Quebec’s AMF, which has its own AI guideline taking effect in 2027.

    Mitey, data editor · Evidence checked

    Canadian Bank AI Index 2026 Q4

    54/ 100

    #7 of 35 · Developing

    • Deployment5
    • Governance5
    • Model risk20
    • Transparency10
    • Talent14

    The editor’s read

    How does Desjardins use AI?

    Regulated by Quebec, with a stricter AI guideline coming

    Desjardins is the one institution in the Index whose main regulator is Quebec’s AMF, so its model-risk evidence is under the AMF’s guideline rather than OSFI E-23, and the AMF’s own AI guideline takes effect for it on May 1, 2027. It applies to any use of AI, requires a risk-based classification of AI systems and includes expectations for the fair treatment of clients.

    Its sourced customer AI is a natural-language phone assistant on AccèsD. We removed Alvie, the in-app tips assistant, because Desjardins’ own 2024 report says its tips used advanced analytics, not AI, at the time. It has a corporate lab at Mila and 11 AI-titled job postings we counted, but no named AI executive and no published AI principles.

    Mitey the yeti inspects a speech bubble through a magnifying glass, checking Desjardins’s sources
    Mitey opened all 8 Desjardins sources, covering 2021 to 2026, and dated each by launch rather than by page.Every row below links to the page it came from.

    How does Desjardins compare with all institutions?

    Desjardins scores 54. The median for all institutions (35) is 11; the Index leader, TD, scores 84.

    Desjardins pillar scores against the all institutions median and the Index leader
    • Desjardins
    • all institutions median
    • Leader (TD)
    • Deployment5 / 3 / 17
    • Governance5 / 0 / 20
    • Model risk20 / 0 / 12
    • Transparency10 / 0 / 15
    • Talent14 / 0 / 20
    Desjardins pillar scores against the all institutions median and the Index leader
    PillarDesjardinsall institutions medianLeader (TD)
    Deployment5317
    Governance5020
    Model risk20012
    Transparency10015
    Talent14020

    Compare Desjardins side by side

    When did Desjardins disclose its AI?

    Entries by launch or first-disclosure year, coloured by who the AI reaches. 1 of 8 reach customers.

    Desjardins AI entries by year and reach2021: 1 Customers use it directly2022: 1 Research & governance2024: 1 Staff tools2025: 1 Staff tools2025: 2 Research & governance2026: 1 Staff tools2026: 1 Research & governance
    Desjardins AI entries by year and reach
    YearCustomers use it directlyRuns on customer accountsStaff toolsResearch & governance
    20211000
    20220001
    20240010
    20250012
    20260011

    Where does each of Desjardins’s 54 points come from?

    Each row is one criterion from the method. A dash means we did not find public evidence, not that the practice is absent.

    How Desjardins’s 54 adds up

    Desjardins meets 8 of 19 criteria. Each pillar is out of 20; the total is their sum.

    How Desjardins’s 54 adds up: inputs, steps and result
    Customer deployment5 / 202 of 7 criteria met
    Governance disclosure5 / 201 of 4 criteria met
    Model-risk readiness20 / 202 of 2 criteria met
    Transparency to customers10 / 201 of 3 criteria met
    AI talent & capacity14 / 202 of 3 criteria met
    Largest gapDiscloses AI or data-science headcount (6 points)
    Total, 2026 Q454 / 100 (Developing)Rank 7 of 35

    Open this comparison →

    Customer deployment5/20

    • Live customer use: Assistants & chatNatural-language phone virtual assistant (AccèsD, built with Google) (2022-06-12)Source 3/3
    • Live customer use: Money insightsNone found in our database0/3
    • Live customer use: Fraud & identityNone found in our database0/3
    • Live customer use: Credit & lendingNone found in our database0/3
    • Live customer use: Wealth & investingNone found in our database0/3
    • Live customer use: Business bankingNone found in our database0/3
    • Customers can use an AI assistant directlyNatural-language phone virtual assistant (AccèsD, built with Google)Source 2/2

    Governance disclosure5/20

    • Publishes responsible-AI principlesNo published AI principles page found (searched desjardins.com, 2025 annual report, 2025 Social and Cooperative Responsibility Report). AR refers to building a 'responsible management framework for AI'.0/5
    • Discloses board or committee oversight of AI2025 Annual Report lists an 'AI and information security' training session for the Audit and Inspection Commission, but no board body with an explicit AI oversight mandate was found.Source 0/5
    • Names AI as a risk in its annual reporting2025 MD&A section 4.1.1 'Main risks and emerging risks' includes an 'Artificial intelligence' risk (bias, reliability, cyber, IP, personal information; risk of lagging adoption).Source 5/5
    • Names an executive accountable for AINo named executive accountable for AI found in 2025 annual report or news searches.0/5

    Model-risk readiness20/20

    • Says its own model-risk framework covers AI/MLNot OSFI E-23: Desjardins applies a model risk framework under the AMF's model-risk guideline that 'encompasses all traditional models, artificial intelligence (AI) and machine learning (ML)'.Source 12/12
    • Names OSFI E-23, or the AMF equivalent for Quebec-regulated institutionsNames the AMF’s model-risk guideline, the Quebec equivalent of E-23 for AMF-regulated institutions.Source 8/8

    Transparency to customers10/20

    • Tells customers when AI or automated decisions are usedPrivacy policy (updated May 25, 2026) says 'we may use AI systems' and that a request 'may be the subject of an automated decision, in which case you will be notified' (e.g. insurance premiums).Source 10/10
    • Signed Canada’s Voluntary Code on generative AINot listed among signatories (archived ISED page 2026-10-23).Source 0/5
    • At least 3 first-party AI disclosures in the last 24 months2 first-party disclosures dated 2024-11 or laterSource 0/5

    AI talent & capacity14/20

    • Has a named AI lab or dedicated AI teamCorporate lab at Mila (2022 agreement); internal AI Centre of Excellence: Mila agreement (2022-08-30, three years) included a 10-workstation Desjardins lab; renewal not verified. AI Centre of Excellence referenced in 2025 media and job postings.Source 8/8
    • Discloses AI or data-science headcountNo count found.0/6
    • Shows active AI hiring11 (Counted by AIBanks.ca on 2026-11-01: postings with AI, machine-learning or data-science terms in the title, from searches for machine learning, artificial intelligence, AI, generative AI and data scientist (plus French terms) on Desjardins’ Workday board (285 total postings), including a manager for its AI Centre of Excellence.)Source 6/6

    What AI reaches Desjardins customers?

    • Natural-language phone virtual assistant (AccèsD, built with Google) (Assistants & chat, 2021)
    Your rights as a Desjardins customer

    All 8 sourced deployments

    2026

    First disclosed

    AI risk governance (AMF guideline engagement, opaque-model supervision)

    Desjardins' 2025 MD&A lists AI as a main/emerging risk, says it is working with the AMF and OSFI on a responsible AI management framework, commented on the AMF's 2025 draft AI guideline, and applies its model-risk framework to AI and ML models.

    Context: Date is the MD&A date.

    Research & governanceGovernance & policyDisclosed in filingsAnnual report: Desjardins Group
    First disclosed

    Generative AI and Microsoft 365 Copilot for staff

    Desjardins' 2025 annual report lists use of generative AI and tools such as Microsoft 365 Copilot among its strategic priorities, and its 2025 responsibility report says mandatory training covered using Microsoft Copilot securely.

    Context: Scale of deployment not disclosed. Date is the MD&A date.

    Staff toolsAssistants & chatDisclosed in filingsAnnual report: Desjardins Group

    2025

    First disclosed

    AI tool for screening job applications

    Desjardins says it uses an AI tool to screen and assess job applications, which it presents as supporting efforts to reduce bias in recruitment.

    Context: Affects job applicants, not members. Date is the reporting-period end; PDF created May 2026. Tool not named.

    Staff toolsStaff & operationsDisclosed in filingsReport / study: Desjardins Group
    First disclosed

    Startup in Residence AI call for startups

    Desjardins' Startup in Residence programme ran Canada-wide calls for startups on artificial intelligence and real-estate technology in 2025, offering grants and possible investment.

    Context: Ecosystem support programme, not a deployment. Date is the reporting-period end.

    Research & governanceResearch & labsLaunchedReport / study: Desjardins Group
    Announced

    AI centre / AI Centre of Excellence

    Desjardins told Finance et Investissement it was setting up an artificial-intelligence centre to pool technology, risk and privacy expertise, with an initial focus on AI applications for employees.

    Context: Desjardins job postings (e.g. builtinmontreal.com job 10565658, closed 2026-08-31) refer to an 'Artificial Intelligence Centre of Excellence Department'.

    Research & governanceResearch & labsAnnouncedNews report: Finance et Investissement

    2024

    Launched

    nCino ML-based financial statement spreading for business lending

    nCino says Desjardins selected its Automated Spreading solution, which uses machine learning to capture financial-statement data and speed commercial credit underwriting, extending a 2022 nCino deployment.

    Context: Vendor-issued. 'Selected' at source date; production rollout not confirmed by Desjardins.

    Staff toolsBusiness bankingLaunchedPress release: nCino (via FF News)

    2022

    Announced

    Strategic AI partnership with Mila (corporate lab)

    Desjardins signed a three-year partnership with Mila covering joint AI and deep-learning research, a 10-workstation Desjardins lab at Mila, and an offer of Mila expertise to Desjardins business members.

    Context: Published on Mila's site. Three-year term from 2022; renewal not verified.

    Research & governanceResearch & labsAnnouncedPress release: Mila

    2021

    LaunchedSource dated In use since December 2021, per La Presse.

    Natural-language phone virtual assistant (AccèsD, built with Google)

    La Presse reports that since December 2021 about half of members calling Desjardins' AccèsD phone service reach a natural-language virtual assistant, built with Google from 5,000 hours of anonymised calls, that handles tasks such as replacing lost cards.

    Context: French-language source. 96.8% word-recognition figure is Desjardins' estimate. Article does not use the word 'IA' explicitly in the quoted passage; it describes natural-language speech understanding.

    Customers use it directlyAssistants & chatLaunchedNews report: La Presse

    Filter these in the database · Download all entries (CSV)

    Questions about Desjardins and AI

    Does Desjardins use AI?

    Yes. Desjardins has 8 sourced AI entries in our database, 1 of them reaching customers, and scores 54 out of 100 (Developing) in the Canadian Bank AI Index, rank 7 of 35. Its strongest pillar is model-risk readiness (20/20); its weakest is governance disclosure (5/20).

    Why does Desjardins score 54 and not higher?

    It did not meet 11 of 19 criteria. The largest are discloses AI or data-science headcount (6 points), publishes responsible-ai principles (5 points), discloses board or committee oversight of AI (5 points). Each can be met with a public, dated source; send one through our tips form.

    Is this a rating of Desjardins as a bank?

    No. The Index scores public evidence about AI: what is deployed for customers, what is disclosed about oversight, and whether customers are told. It does not rate service, fees, safety or whether the AI is accurate or fair.

    Change log

    1. Launch dates corrected from each entry’s own source text: TD mortgage/HELOC pre-approval (August 2023), TD Insurance term life (March 2023), TD nudges (November 2020), TD U.S. Virtual Assistant (April 2020), TD Insurance FRISS (June 2021), Desjardins phone assistant (December 2021), Scotiabank predictive prompts (end of 2025), RBC ATOM (in use since 2023) and three CIBC tools (2025). A test now fails if an entry’s text names an earlier launch than its launch date.

    2. Removed three rows whose sources do not say AI is used: Desjardins Alvie (its own report says analytics, not AI, for now), ATB’s 2026 virtual assistant, and RBC Genesis. Added AI job-posting counts for CIBC (12) and Desjardins (11); Scotiabank and National Bank job boards could not be counted and score 0 on that criterion until they can.

    3. Scored in the 2026 Q4 edition with 8 sourced entries; every row verified on the date shown with it.

    Sources (9)

    1. At least 3 first-party AI disclosures in the last 24 months · Desjardins
    2. 2025 Social and Cooperative Responsibility Report · Desjardins Group · 2025-12-31
    3. Desjardins : une transformation numérique portée par l'écoute de ses membres · Finance et Investissement · 2025-05-06
    4. Desjardins selects nCino's single platform to automate loan origination process through machine learning · nCino (via FF News) · 2024-03-21
    5. Has a named AI lab or dedicated AI team · Desjardins
    6. Customers can use an AI assistant directly · Desjardins
    7. Tells customers when AI or automated decisions are used · Desjardins
    8. Signed Canada’s Voluntary Code on generative AI · Desjardins
    9. Shows active AI hiring · Desjardins

    Explainers for these uses

    Scores reflect public evidence found by 2026-11-01. They are not a rating of service, safety or financial strength, and not a recommendation. Institutions can send corrections to hello@mitey.ca.

    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.