What problem does it solve?
Every investment recommendation to a retail client needs a documented suitability assessment: does the product or portfolio match the client's objectives, horizon, risk tolerance, capacity for loss, knowledge and experience, and sustainability preferences, and why. Under MiFID II the client receives a written statement of suitability, and comparable suitability or best interest duties apply in other major markets.
In practice the rationale is written by hand from a fact find, a risk profile and product data held in different systems. Quality varies by advisor, reports are long and generic, and supervisors find gaps only when sampling after the fact. Circumstances also change: a report that was right at the time of advice says nothing about whether the portfolio still fits a year later.
How does it work?
- Assemble the facts. The assistant pulls the client's profile (objectives, horizon, risk tolerance, capacity for loss, knowledge and experience, preferences) and the proposed products or portfolio with their risk and cost data.
- Run the rules. A deterministic rules engine applies the firm's suitability and product governance rules (risk class limits, target market, concentration, complexity) and returns pass, fail or refer, with reasons.
- Reason about the gaps. For referrals, the model explains the mismatch in plain language and suggests what information is missing or what the advisor should consider.
- Draft the report. A suitability rationale is drafted in the firm's template, quoting the client's own stated objectives and the rule results, never inventing facts.
- Advisor and supervisor confirm. The advisor edits and confirms the report; referrals and hard fails go to a supervisor. Everything is logged for audit.
- Audience
- Employee facing
- Autonomy
- Copilot
- Adoption
- Emerging
- Channels
- Internal tools
What is it worth?
Benchmarks are computed from the public deployments below: one data point per organization per KPI, with who made each claim.
No public deployment has disclosed a measurable outcome yet.
Value drivers: Compliance quality, Employee productivity, Risk and loss reduction, Customer experience.
Indicative value
A wealth manager with 500 advisors giving regulated advice
USD 1 million to USD 7.5 million
Value of advisor time released from suitability write up per year
How this is calculated
Formula: advisors * recommendations * minutesSaved / 60 * hourlyCost. The low scenario uses every low input, the high scenario every high input.
| Input | Low | High | Basis |
|---|---|---|---|
| Advisors giving regulated advice advisors, advisors | 500 | 500 | The reference firm. |
| Recommendations with a suitability report per advisor per year recommendations, reports per advisor per year | 100 | 200 | Editorial assumption, replace with your own advice volumes. |
| Minutes saved per suitability report minutesSaved, minutes per report | 15 | 30 | Editorial assumption, replace with your own. No public source on this page states a time saving for AI drafted suitability reports. |
| Fully loaded advisor cost per hour hourlyCost, USD per hour | 80 | 150 | Editorial assumption, replace with your own fully loaded cost. |
What it leaves out: Productivity only. The main value is consistent, complete suitability records and fewer unsuitable recommendations, which this figure does not price, and it leaves out rules engine and review costs.
Who already uses it?
2 public deployments, strongest evidence first. Grades: A regulator or audit, B the organization itself, C vendor case study, D anonymous or estimate.
Vanguard
United States · Wealth and asset management · 2020
Vanguard Digital Advisor is an all digital advice service that gathers a client's goals, time horizon and risk tolerance, uses an algorithm to build a portfolio for them, and keeps it on track with ongoing monitoring. It rebalances when a portfolio drifts more than 5% from the recommended allocation and adjusts holdings when a client adds goals. Vanguard discloses limits of the automated assessment, for example that it does not assess the suitability of selling existing holdings, and it labels its projections and goal forecasts as hypothetical and educational, not guarantees. The service was live by 2020: a Vanguard page archived in December 2020 refers to Digital Advisor clients who enrolled before 1 October 2020.
No outcome disclosed.
Morgan Stanley
United States · Wealth and asset management · 2025
BlackRock announced on 2 October 2025 that Morgan Stanley Wealth Management's Portfolio Risk Platform would be the first to implement Auto Commentary, a generative AI feature of Aladdin Wealth, with advisors in the U.S. getting access from October. The tool combines Aladdin risk analytics, the firm's Chief Investment Office outlook and the client's holdings and investment preferences to draft concise insights for the advisor, highlighting issues such as overweights or misalignment with the client's objectives or the firm's market view. Trade press describes the output as bullet point insights inside a template, not full scripts or emails, so it supports the advisor's conversation rather than producing a finished client report. No outcome figures were published.
No outcome disclosed.
How do you implement it?
A model agnostic playbook: what to prepare, the order to build in, and what goes wrong.
Data you need
- Client fact find and risk profile with dates of last update
- Product data such as risk class, complexity, costs and target market
- Codified suitability and product governance rules per market
- Approved suitability report templates and wording
Systems to integrate
- Advice and portfolio management platform
- Product governance and target market database
- CRM for client profile and records
- Document generation and client delivery
Complexity: High
This is a regulated control. It needs codified suitability and product governance rules, clean client profile data, a validated risk profiling method, compliance sign off on the report template and a full audit trail.
- 1
Codify the rules before adding a model
Write the firm's suitability and target market rules as deterministic checks with clear outcomes. The model explains and drafts; it does not decide hard fails.
- 2
Draft from facts, not from memory
The report may only cite the client's recorded profile, the rule results and product data. Anything the model cannot trace to a source is left out or flagged.
- 3
Pilot on one advice type
Start with a common, simple advice type (for example a model portfolio recommendation) and compare AI drafted reports with manual ones in supervisory review.
- 4
Add ongoing suitability monitoring
Once point in time assessments work, rerun the checks when markets or client circumstances change and flag portfolios that no longer fit for advisor review.
- 5
Keep compliance in the loop
Compliance approves templates, rule changes and model changes, and samples reports every month.
Guardrails
- Hard suitability failures decided by deterministic rules, never overridden by the model
- Reports cite only recorded client facts, rule results and product data
- Advisor confirmation of every report, supervisor review of referrals and overrides
- Full audit log of inputs, rule outcomes, drafts and edits
- No use of protected characteristics in suitability reasoning
KPIs to instrument
- Time to a confirmed suitability report
- Supervisory findings per hundred reports, before and after
- Share of assessments referred or failed, with reasons
- Advisor edit rate on drafted rationales
- Portfolios flagged by ongoing monitoring and resolved
Human in the loop
The advisor confirms each assessment and report and remains responsible for the recommendation; supervisors decide referrals and overrides; compliance owns the rules and templates and samples output.
Common failure modes
- Model overrules the rule
- A fluent rationale justifies a product that failed a hard check. Keep the verdict in the rules engine and block contradictory drafts.
- Boilerplate reports
- Every report reads the same and does not reflect the client's own words. Require quotes from the fact find and supervise for generic text.
- Stale profiles
- The assessment uses a risk profile that is years old. Check profile dates and require an update before advice.
- Hidden bias
- Reasoning uses proxies such as age or nationality inappropriately. Test outputs across segments and restrict inputs.
What are the risks and rules?
EU AI Act
Depends on design
Investment suitability assessment is not listed in Annex III, so the tier depends on design. It becomes high risk where the same system assesses creditworthiness, for example for lending against a portfolio (Annex III point 5(b)). MiFID II suitability duties apply regardless of the AI Act tier.
Rules that apply
Guidance
- Guidelines on certain aspects of the MiFID II suitability requirements (European Securities and Markets Authority, Europe). The current guidelines (ESMA35-43-3172, 2022), which replaced the 2018 version (ESMA35-43-1163). They set out how firms collect client information, including sustainability preferences, assess suitability and document it, with specific points on automated advice.
- ESMA public statement on the use of AI in the provision of retail investment services (European Securities and Markets Authority, Europe). Says firms that use AI in investment advice and portfolio management must apply heightened vigilance and diligence, particularly in ensuring the suitability of services and financial instruments for each client.
- Artificial Intelligence Model Risk Management (information paper) (Monetary Authority of Singapore, Asia Pacific). Good practices observed in a 2024 MAS thematic review of banks' AI and generative AI model risk management, covering governance and oversight, risk management systems and processes, and development and deployment.
Controls to put in place
- Rules engine and model inventoried with owners in compliance and the business
- Version control and approval for rules, templates and prompts
- Audit trail per assessment kept for the regulatory retention period
- Monthly supervisory sampling with documented findings
- Fairness testing of drafted rationales across client segments
Frequently asked questions
- Can AI decide whether an investment is suitable?
- It should not make the verdict on its own. Keep hard rules in a deterministic engine, let the model explain and draft, and have the advisor confirm. ESMA expects heightened diligence on suitability when AI is used in advice.
- Is anyone doing this at scale yet?
- Automated digital advice services such as Vanguard Digital Advisor already use an algorithm to build portfolios from a client's goals, time horizon and risk tolerance, and since October 2025 Morgan Stanley's advisors get AI drafted talking points that flag misalignment with a client's objectives. Public, named deployments of generative AI that write full suitability reports are still rare, so treat vendor claims with care.
- What about checking suitability after the sale?
- That is where agentic monitoring helps: rerunning the checks when markets or client circumstances change and flagging portfolios for review. It links closely to drift monitoring.
How to cite this page
Blits.ai AI Use Case Library, "AI assistant for investment suitability assessment and reports", last verified 26 September 2026, https://www.blits.ai/ai-use-cases/suitability-assessment-assistant. Licensed under CC BY 4.0. Method: how we verify use cases.
Changelog
- 27 September 2026: First published