Suitability is a judgement about a person, not a portfolio. That single fact makes wealth advisers the real control once AI agents prepare review packs and rebalancing proposals: the adviser decides whether a proposal suits this client, now, given everything they know about them, and answers for it.
What does an agent prepare for a wealth adviser?
The analysis and the paperwork. At depth 1 the agent reads holdings, transactions, the client's risk profile, stated objectives and house views. At depth 2 it drafts: a portfolio review showing performance and drift, a rebalancing proposal against the model, a suitability report in the firm's format, and meeting notes from a recorded call. Routine administration, such as scheduling reviews or sending standard statements, may be certified at depth 3.
Executing trades is depth 4: money moves and the market will not reverse it. Most firms keep trade execution held for an adviser's approval, and some keep any change to a client's portfolio entirely with people.
Why is suitability a human judgement?
Because the inputs are partly human. A client's risk tolerance, objectives and circumstances are recorded in a profile, but that profile is mostly declared: the client said it, at a point in time. Under the Pryme Context Protocol, declared facts are labelled as such, distinct from sealed data such as holdings from the custody system.
A client who said "balanced" three years ago may now be retired, newly widowed or worried about a child's finances. The agent's proposal will be consistent with the profile. Whether the profile is still true is the adviser's question. Advisers who read the provenance and date of every client fact before reading the proposal are doing the work that makes advice suitable.
Where do agent proposals go wrong?
Agent drafts in wealth management tend to fail in predictable places:
- Stale client information. Objectives or circumstances that changed but were never updated.
- Concentration the model ignores. Holdings outside the managed portfolio, such as a business or property, that change the real risk.
- Tax effects. Rebalancing that is sensible on paper but triggers gains the client would rather defer.
- Vulnerability. Signs of cognitive decline, bereavement or pressure from family that a proposal cannot see.
- Tone. A technically correct review letter that alarms a nervous client.
Each is a reason the adviser's review cannot be a formality.
There is also a subtler failure. Agents draft to the firm's house view and model portfolios, which is what they were asked to do. Over time, every client's proposal starts to look the same, because the agent is consistent and the adviser stops pushing back. Suitability is meant to be individual. An adviser who notices that their last twenty proposals were nearly identical, for clients with very different lives, has found the point where consistency has quietly replaced judgement.
How does the meeting with the client change?
It gets more time, if the adviser protects it. The hours an agent saves on preparing packs and drafting reports should move into the conversation: asking what has changed, testing whether the client still means what their profile says, and explaining the proposal in plain words. The agent can prepare a list of questions based on gaps in the record. Only the adviser can notice the hesitation in a client's voice when the subject of their children's finances comes up, and decide what it means for the advice.
What should a wealth adviser never hand over?
The recommendation itself. An agent can propose; the adviser recommends, because the recommendation carries their professional responsibility. Trade execution, changes to the client's risk profile, and anything involving a client who shows signs of vulnerability should stay with the adviser. So should conversations about losses, which need a person.
The record supports this. Every verdict and every human decision is filed in an append-only record. If a client later complains that advice was unsuitable, the record shows what the agent proposed, what the adviser approved, and what the adviser knew at the time. That protects careful advisers and exposes rubber-stamping.
What does a wealth adviser need to learn?
How the agent decides. Depth, so they know whether a trade was executed or only proposed. The minimum chain: five bounds apply to every action, the lowest wins, and the verdict names it. Verdicts: held for a named person, refused as outside the agent's bounds, or withheld at the platform floor for every agent. Provenance, especially for client-declared facts. And certificates: they lapse after 90 days, and a run that clears no depth drops the agent to depth 1 the same day, so review packs the agent prepared may stop arriving complete.
They also need a review routine that starts with the client, not the portfolio. The detail is in how wealth advisers review an AI agent's decisions.
How do wealth advisers compare with other advisers?
HR advisers and wealth advisers both work with people at important moments and both review agent drafts that are consistent with the record but may miss the person. Payments operations staff, by contrast, deal with the money movement after the decision is made. See HR advisers as the real control and payments operations staff as the real control, and the overview in AI agent skills by role.
Markets differ in regulation and client expectations. The briefs on wealth advisers in Germany and Kenya cover two very different settings. Pryme Intelligence Academy sets module tests in the learner's own industry and role, so advisers practise on wealth cases.
Questions people ask
Can an agent rebalance portfolios automatically?
For discretionary mandates within tight bounds, a firm might certify routine rebalancing at depth 4 after strong evidence. Many keep every trade with an adviser. Advisory relationships, where the client decides, should not have automatic trades.
How should an adviser use agent-drafted meeting notes?
Read and correct them before they go on file. Notes become evidence of what the client said and what was advised, so errors matter later.
Does the agent reduce the time spent on suitability reports?
It reduces drafting time. The time saved should go into the client conversation and the review of what the agent assumed, not into more clients per adviser at the same depth of attention.