Suitability under the client-focused rules, signs that older clients are being exploited, conflicts and source of wealth fill a Canadian wealth compliance team's week. AI agents can review account files and trades at scale, spot exploitation signals and draft source of wealth assessments. Supervisors and compliance officers decide. Provincial securities regulators, OSFI for bank-owned firms and PIPEDA apply.
What do wealth compliance teams in Canada spend their time on?
Suitability work changed when the client-focused reforms took effect. Firms must know their clients in depth, know the products they offer, put the client's interest first when determining suitability, and address material conflicts in the client's best interest. Compliance tests whether account files support each recommendation, whether the product shelf was understood, and whether conflicts such as proprietary products were handled.
Older and vulnerable clients are a growing strand. Firms must ask clients for a trusted contact person and may place temporary holds where they suspect financial exploitation or diminished capacity. Compliance reviews whether those tools were used properly, and whether signs were missed.
Anti-money laundering work runs throughout: identifying clients, verifying source of funds and wealth for higher-risk relationships, monitoring transactions and reporting to the federal financial intelligence unit.
Where can AI agents help?
- Suitability file review (depth 1). The agent reads the know-your-client form, the investment policy statement and the trades, and flags mismatches: a conservative client holding concentrated positions, a time horizon that does not fit the products, or a proprietary fund chosen with no documented comparison.
- Exploitation signals (depth 1). The agent watches older clients' accounts for patterns such as sudden large withdrawals to new payees, a new person giving instructions, or address changes followed by transfers. It flags the account and checks whether a trusted contact is on file.
- Source of wealth assessment (depth 2). The agent reads the client's declared source of wealth and the supporting documents, tests whether they fit together, and drafts an assessment for an analyst. Our playbook on reviewing source of wealth evidence with an AI agent covers each step.
- Case creation (depth 2). Each flag opens a draft case in your case management system with the evidence attached, so supervisors work from one queue. See our guide on connecting an AI agent to your case management system.
The exploitation use needs care with sources. A transfer record is sealed evidence from the custody system. A phone note saying the client sounded confused is a person's observation, declared and unverified. Both matter, but the agent must label each, so the supervisor knows what the flag rests on. The guide to AI compliance agents explains why this labelling decides whether a finding holds.
What stays with people?
Placing a temporary hold on a client's account is a serious step. It interferes with the client's control of their money and must rest on a reasonable belief. The agent can surface signs; a trained supervisor speaks to the client or trusted contact and decides.
Suitability conclusions, conflict findings and decisions to discipline an adviser stay with named people. So does the decision to file a suspicious transaction report, which carries legal duties and must be made by people trained to make it.
Client acceptance or exit on source of wealth grounds is human. Wealthy clients often have complex histories, and the judgement on whether a gap is innocent needs context.
Which rules apply in Canada?
Provincial securities regulators set and enforce securities rules for advisers and dealers, including the client-focused requirements on know-your-client, know-your-product, suitability and conflicts, and the rules on trusted contacts and temporary holds. A national self-regulatory organisation oversees investment and mutual fund dealers.
Where the wealth business sits within a federally regulated bank, OSFI supervises the group, and OSFI Guideline E-23 on model risk management sets expectations for governing models, which reach an agent used in material processes. FCAC oversees consumer protection obligations of federally regulated banks.
PIPEDA governs personal information in the private sector in most provinces; Quebec and some other provinces have their own laws. Anti-money laundering duties follow the FATF Recommendations. There is no federal AI act in force; our guide to how AI agents are regulated in Canada covers the frameworks that fill the gap.
This is an explanation of the rules, not legal advice; confirm how they apply to you with counsel.
How to start in the first 90 days
Days 1 to 30: choose suitability file review for one product type, such as accounts holding proprietary funds. Register the agent in your model inventory if you are bank-owned, and name the supervisor who owns findings.
Days 31 to 60: replay the agent on files your compliance team already reviewed. Read every disagreement; some will show where your own checklist is unclear.
Days 61 to 90: certify at depth 1 and start exploitation signals on clients over a set age, with supervisors reviewing every flag. Indian wealth firms face different priorities; see our brief on compliance agents in Indian wealth management.
Questions people ask
Can an agent place a temporary hold?
No. A hold requires a reasonable belief formed by a person and a conversation with the client or trusted contact. The agent flags the signs.
Will the agent judge clients by age?
Age can be a factor in deciding which accounts to watch, but flags should rest on behaviour. Review the agent's flags for fairness so older clients are not treated as presumed vulnerable.
Does Quebec require anything extra?
Quebec's privacy law adds duties, including on automated decisions and transfers outside the province. Map these before the agent reads Quebec clients' data.