Financial Services Spoke

AI in Wealth Management and Fiduciary Risk

The advisor-level pull on AI is the strongest in financial services. The investor-trust gap and the fiduciary framework are tightest in wealth management.

The Advisor Productivity Pull

Wealth management has the strongest advisor-level productivity pressure on AI adoption

42%
Of bank advisors currently use AI (Cerulli 2025)
77%
Projected bank advisor AI use within two years (Cerulli 2025)
23%
AI use growth since 2023 across 42 RIAs and B-Ds with $6T AUM (F2 Strategy)
25–50%
Productivity gain on client letters, market commentary, meeting prep

The Investor Trust Gap

Cerulli's 2025 investor research found only 38% of affluent investors are at least somewhat comfortable with AI in financial advice — essentially flat from 39% in 2024. The figure drops to 16% among investors aged 70+.

The strategic implication — advisors using AI need to be cautious about disclosure. The competitive advantage of AI-assisted advice is real; the client-relationship risk if disclosure is mishandled is also real.

Fiduciary Duty Under the Investment Advisers Act

AI-assisted advisers owe the same fiduciary obligations as advisers using only human judgment — there is no AI carve-out

Pillar 1

Duty of Loyalty

The adviser must eliminate or fully disclose conflicts of interest. If the AI tool's design creates incentives the adviser does not disclose, or if the adviser receives compensation related to specific AI-recommended products, the conflict is in scope.

Why it matters: AI-tool design choices become disclosure questions.

Pillar 2

Duty of Care

The adviser must provide investment advice that is in the client's best interest based on the client's investment profile. Recommendations driven by AI must still meet suitability and best-interest standards.

Why it matters: AI does not lower the suitability bar — it adds a documentation layer.

Robo-Advisor Enforcement as the Operational Template

The SEC's robo-advisor enforcement line tells you how AI-assisted advice enforcement is likely to unfold

Settled

Betterment

Settled fiduciary-duty allegations for failure to disclose material service changes.

Why it matters: Disclosure adequacy, not algorithmic defect, was the enforcement lever.

Settled

Wahed Invest

Settled allegations relating to misrepresentations about advisory services.

Settled

SoFi

Settled allegations relating to robo-advisor practices.

Settled

Schwab

Settled robo-advisor allegations.

Mar 2026

Ally Invest Advisors

Administrative order for a six-year undisclosed conflict in its Cash-Enhanced robo-advisor accounts.

Why it matters: Most recent — confirms enforcement remains active on disclosure adequacy.

Pattern

Disclosure Is the Lever

Enforcement focuses on conflicts of interest and disclosure adequacy, not on whether the algorithm itself was technically defective.

Why it matters: AI-assisted advice enforcement will follow the same shape.

ERISA Fiduciary Risk — Currently Forward-Looking

The DOL's 2024 Retirement Security Rule, which would have expanded the definition of ERISA fiduciary, was vacated by federal courts and formally removed by the DOL on March 18, 2026. The 1975 five-part test for ERISA fiduciary status remains in place.

No named ERISA case turning specifically on AI-generated advice has been filed as of mid-2026. This is an emerging-risk area — as AI-assisted advice penetrates 401(k) plan-level advice, retirement account guidance, and rollover recommendations, ERISA fiduciary exposure will follow.

Governance Controls Every Wealth Management AI Deployment Needs

Seven controls that translate fiduciary duty and Reg BI into operational practice

1

Form ADV and engagement letter review

Disclosures match operational reality. The Delphia question — does what you say match what you do?

2

Reg BI documentation

Every AI-assisted recommendation has a documented best-interest basis.

3

Supervision under Rule 3110

AI-assisted communications and recommendations flow through supervisory review.

4

Recordkeeping under 17a-4 and 4511

AI prompts and responses for client-affecting interactions retained per the retention rules.

5

MNPI / private wealth controls

Where the wealth management business intersects with corporate clients or insiders, the same MNPI risks apply as in research.

6

Conflict of interest review

AI-tool fee structures, model-product bundling, and revenue-sharing arrangements scrutinized for conflicts.

7

Client disclosure

Material AI use disclosed at the appropriate level — Form ADV, engagement letter, or client communication.

AI in Wealth Management — FAQ

How widely do wealth advisors use AI today?

Cerulli's 2025 research found 42% of bank advisors currently use AI in their practice, projected to rise to 77% within two years. F2 Strategy's 2025 wealth-management survey reported 23% growth in AI use since 2023 across 42 RIAs and broker-dealers representing $6 trillion AUM.

Does using AI change an adviser's fiduciary duty under the Investment Advisers Act?

No. AI-assisted advisers owe the same duty of loyalty (conflicts disclosure) and duty of care (suitable, best-interest recommendations) as advisers using only human judgment. There is no AI carve-out. The substantive obligation is identical; the documentation and disclosure obligations adapt to the AI's role.

Has there been an SEC enforcement action involving AI-generated investment advice and fiduciary breach?

No named case yet turns squarely on AI-generated advice and fiduciary breach. The closest analogues are the robo-advisor enforcement actions against Betterment, Wahed Invest, SoFi, Schwab, and Ally Invest Advisors (March 2026 order) — focused on disclosure adequacy and conflicts of interest. The pattern is likely to carry into AI-assisted advice enforcement.

Was the DOL Retirement Security Rule about AI?

No, but its fate affects how AI-assisted advice on retirement accounts will be regulated. The 2024 Retirement Security Rule, which would have expanded ERISA fiduciary status, was vacated by courts and formally removed by the DOL on March 18, 2026. The 1975 five-part test for ERISA fiduciary status remains in place — meaning AI-assisted advice on retirement accounts is governed by long-standing ERISA principles, not a new framework.

Do advisers need to disclose AI use to clients?

It depends on materiality. Form ADV must disclose AI use where material to the advisory relationship (the Delphia / Global Predictions framework explicitly targets misrepresentation in AI disclosures). Engagement letters generally should address AI use where AI is integral to the service. The conservative position is to disclose at the Form ADV and engagement-letter level whenever AI materially affects service delivery.

Build a Fiduciary-Defensible AI Program in Wealth Management

Free Shadow AI Assessment audits your Form ADV disclosures, your Reg BI documentation, your conflict-of-interest review, and your client-disclosure framework.