Where Should an RIA Start With AI?
Dr. Leigh Coney
Founder, WorkWise Solutions
November 5, 2026
6 min read
TLDR: An RIA should start with AI where the work is already written down and a person already checks it: meeting notes and follow-up tasks, first drafts of client service email, onboarding paperwork, and answers from the firm's own procedures. Pick two for a 90-day pilot on a business plan, keep what you send to clients as a record, and leave investment advice and anything sent to a client without review out of scope. WorkWise Solutions runs this first step as a $12,500 AI Readiness Sprint for firms of up to 20 people.
I run the AI Readiness Sprint through WorkWise Solutions: $12,500 flat for firms of up to 20 people, one to two weeks, ending in ranked use cases, guardrails and a 90-day plan your team can manage. Firms that want ongoing help then keep me on as their AI advisor from $10,000 a month on the AI Operating Partner retainer.
Table of Contents
1. The Short Answer
An RIA should start with AI where the work is already written down and a person already checks it: meeting notes and follow-up tasks, first drafts of client service email, onboarding paperwork, and answers from the firm's own procedures. Pick two for a 90-day pilot on a business plan, keep what you send to clients as a record, and leave investment advice and anything sent to a client without review out of scope.
A business plan is the paid company version of an AI tool, such as Claude Team or ChatGPT Business. Anthropic and OpenAI both say they do not train their models on business-plan content by default.
In Schwab's RIA & AI Research Study (January 2026, 533 advisors who custody at Schwab), 63% of advisors used AI, and 70% of those users were "individuals exploring tools on their own."
A practical AI strategy turns that solo use into firm practice: two jobs, written guardrails, one owner and 90 days.
2. The Five Areas, Ranked
The five areas, ranked for a firm of 10 to 25 people:
| Area | First job to hand AI | Value | Effort | Guardrail |
|---|---|---|---|---|
| 1. Recurring administrative work | Meeting notes, follow-up tasks and scheduling emails | High: 71% of advisors using AI already use it for notes, scheduling and email drafts | Low | The advisor reads each summary before it goes into the CRM |
| 2. Client service | Meeting prep and the first draft of the follow-up email | High: 39% of AI users draft client communications; some firms cite an hour or more saved on meeting prep | Low | The advisor approves every email; the sent copy is kept |
| 3. Onboarding | A new client's statements and forms turned into one intake summary | High: the typical team needs 18 staff hours and 45 days to complete and implement a first plan | Medium | A second person checks every figure against its statement |
| 4. Internal knowledge | Answers to staff questions from the firm's own procedures | Medium: 22% of AI users create internal training materials; 7% of firms have built a prompt library (shared, tested instructions) | Low | No client data; the CCO approves the source procedures |
| 5. Operations | A weekly summary of open service requests, from reports your team already exports | Medium at first: 28% of AI users analyze firm data; 26% automate processes | Medium | Read-only access: AI can look but not change anything |
The top rows share one test: the work is already written down, and a person already checks it.
Operations comes last for a first pilot, even though advisors in the T3/Inside Information Software Survey 2026 (March 2026, 2,906 responses) expect the most from back-office efficiency, 7.72 out of 10. Its bigger wins need AI connected to your CRM, which is easier to approve once the first pilots work.
Two jobs stay off the list: investment advice, and a chatbot that answers clients. The same advisors expect the least from client-facing chatbots, 4.37 out of 10.
3. What a 10 to 25 Person Firm Does First
If you have been asked for a practical AI strategy with appropriate guardrails and an implementation plan that is manageable for your team, start with two pilots.
Pilot one is the meeting cycle, rows 1 and 2: prep, notes, follow-up tasks and the follow-up email, for two or three advisors. The Kitces Report found that senior advisors "spend more than one hour of prep and follow-up for every one hour of client meetings."
Pilot two is onboarding, row 3, for the operations team. A firm that wants no client data in its first pilots takes internal knowledge instead.
The COO or chief of staff owns the plan: picks the pilots, books the people and counts the hours. The CCO signs the rules. Rule 206(4)-7, the SEC's compliance program rule, requires written policies and procedures "reasonably designed to prevent violation," reviewed at least once a year, with a chief compliance officer in charge, so AI use belongs in them.
The tool is a separate choice: What Is the Best AI for Financial Advisors? compares Claude, ChatGPT and, for firms on Microsoft 365, Microsoft 365 Copilot.
4. Guardrails in Plain Words
Appropriate guardrails answer three questions for each area: what goes in, who checks it and what the firm keeps. The Kitces AdvisorTech Study (September 2025, 706 responses) names the same three where it calls compliance "a major barrier."
What goes in
Client data goes only into the firm's business plan, never a personal account. Meeting and onboarding work needs client details, but no summary needs a Social Security or account number. Internal knowledge needs no client data at all.
Who checks it
The advisor reads every meeting summary and approves every client email. In operations, a second person checks each onboarding figure, and each weekly summary against its source reports.
The CCO approves the procedures before staff rely on answers from them, because AI can state a wrong answer with confidence.
What is kept
Rule 204-2, the SEC's books-and-records rule, requires an adviser to keep copies of the written communications it sends about its recommendations and advice for "not less than five years." An AI draft that goes to a client is kept like any other client email.
Decide in writing, with your CCO, where meeting and onboarding summaries are kept. Is Claude Compliant for SEC and FINRA? covers the SEC and FINRA record rules in full.
5. A 90-Day Plan Your Team Can Manage
The plan runs on each week's real work, so nobody stops a day job for it.
- Days 1 to 15, rules and tool. Pick the two pilots and the tool, and have the CCO sign a one-page AI use rule. Anyone on a personal account moves to the firm's plan.
- Days 15 to 45, two pilots. Two or three people per pilot use AI on real meetings and new clients, and log the hours each job takes.
- Days 45 to 75, write down what works. Save the instructions that worked in the prompt library, and add the review steps to your procedures.
- Days 75 to 90, decide what to add. Keep, fix or drop each pilot, pick the next area from the table, and add AI to the annual compliance review.
At day 90 you have two measured jobs, signed rules and a library the next hire can use.
6. Where to Start
- List who already uses AI, for which jobs and on which accounts.
- Count today's hours for meeting prep and follow-up, and for each new client's paperwork.
- Pick two jobs from the top three rows of the table.
- Write one page of rules with your CCO before anyone starts.
- Name the owner and put days 15, 45, 75 and 90 on the calendar.
WorkWise Solutions, which publishes this guide, runs the first step as the AI Readiness Sprint: $12,500 flat for firms of up to 20 people, one to two weeks, ending in a prioritized use-case map, a vendor-neutral platform decision and a 90-day plan. Firms of 20 or more take the Comprehensive Discovery Sprint, $30,000. Once the pilots run, the AI Operating Partner retainer keeps an advisor on call from $10,000 a month for weekly office hours, one build at a time and a monthly usage review, with a 3-month minimum; base prices cover firms of up to 25 people.
WorkWise sells no licenses and takes no referral fees. Not a fit if someone inside already owns AI and has the hours to run the 90 days alone.
"Compliance also remains a major barrier. Firms need clarity on how client data is handled, whether AI tools are archivable, and how output is reviewed and approved to avoid hallucinations"
The Kitces Report, AdvisorTech Study, Vol. 1, 2025, based on 706 responses (September 2025)
- •An RIA should start with AI where the work is already written down and a person already checks it: meeting notes and follow-up tasks, first drafts of client service email, onboarding paperwork and answers from the firm's own procedures.
- •In Schwab's RIA & AI Research Study (January 2026, 533 advisors who custody at Schwab), 71% of advisors using AI used it for notes, scheduling and email drafts, and 39% used it to draft client communications.
- •A firm of 10 to 25 people should run two pilots over 90 days, with the COO or chief of staff owning the plan and the CCO signing the rules on what data goes in, who checks the output and what is kept.
- •Rule 204-2, the SEC's books-and-records rule, requires an adviser to keep copies of the written communications it sends about its recommendations and advice for not less than five years, so an AI draft sent to a client is kept like any other client email.
- •Advisors in the T3/Inside Information Software Survey 2026 (2,906 responses) expect the most from back-office efficiency, 7.72 out of 10, and the least from client-facing chatbots, 4.37, which is one reason a client chatbot stays out of the first pilots.
Frequently Asked Questions
What is the first AI use case for an RIA?
A good first AI use case for an RIA is the meeting cycle: meeting prep, notes, follow-up tasks and the first draft of the follow-up email, each checked by the advisor. Many advisors already use AI this way: in a January 2026 Schwab study of 533 advisors who custody at Schwab, 71% of those using AI used it for notes, scheduling and email drafts. Run it on a business plan, keep the emails you send as records, and leave investment advice out of scope.
How long does it take an RIA to roll out AI?
An RIA of 10 to 25 people can roll out its first AI uses in about 90 days: two weeks to set the rules and pick the tool, a month on two pilot jobs, a month to write down what works, and two weeks to decide what to add. Wider use takes longer, because each new job needs its own reviewer and written steps. A January 2026 Schwab study of 533 advisors found most in the early stages, with 30% experimenting and 10% strategically integrating AI.
Who helps an RIA build a practical AI strategy?
An RIA builds a practical AI strategy around two people inside: the COO or chief of staff, who owns the plan, and the CCO, who signs the rules. Outside help fits when nobody inside has the hours. WorkWise Solutions, which publishes this guide, is a boutique AI advisor for RIAs: its AI Readiness Sprint costs $12,500 for firms of up to 20 people and ends with ranked use cases, a platform decision and a 90-day plan, and its AI Operating Partner retainer keeps an advisor on call from $10,000 a month.
Related Guides & Articles
AI Consulting for RIAs
A boutique AI advisor for RIAs: the use cases worth doing first, the guardrails your CCO will ask for and an implementation plan, from $10,000 a month.
What Is the Best AI for Financial Advisors?
Claude, ChatGPT or Microsoft 365 Copilot for an RIA, compared on plans, prices, records and fit.
Where Should a PE Firm Start With AI?
The same question for private equity: what makes a good first workflow, and how to prove it with one team and one number.
Claude for Financial Advisors for Small RIAs
Team or Enterprise for an RIA under 20 people: what each costs at 3, 5 and 8 people, and what Team gives up on records.
Asked to build a practical AI strategy for your RIA?
Tell me your headcount, the systems your team runs and where the hours go each week, and I'll tell you which two jobs to start with and what the first 90 days cost. The AI Readiness Sprint costs $12,500 flat for firms of up to 20 people; firms of 20 or more start with the $30,000 Comprehensive Discovery Sprint.