The AI tools for financial advisors that firms actually use in 2026 are mostly back-office helpers: meeting note-takers, tax-return readers, and assistants that search research and draft emails. They save time on paperwork rather than pick investments, and every output still falls under the same securities rules as work done by hand, so an advisor has to review, keep records of, and supervise what the AI produces.
This guide maps the main categories with named examples and published prices, then covers the regulatory side and what clients can reasonably ask their advisor. It is written for independent advisors weighing tools and for clients who want to know what happens to their data.
What advisors use AI for
FINRA surveyed its member firms and found they are starting with efficiency gains in internal processes and information retrieval. The single most common use is summarization and information extraction: condensing long documents and pulling out the key facts (FINRA 2026 Regulatory Oversight Report). Its catalog of observed uses also includes drafting content, chat-style question answering, translation, classification, data transformation and financial modeling (FINRA).
In practice, that breaks down into five kinds of tool.
| Category | What it does | Examples | Published price |
|---|---|---|---|
| Meeting note-takers | Capture client meetings, write notes, draft follow-up emails, push tasks into the CRM | Jump, Zocks, Morgan Stanley Debrief (in-house) | Zocks from $67 per user a month billed annually, $80 monthly (Zocks); Jump’s core plan $100 per advisor a month (Jump) |
| Tax return analysis | Read a scanned 1040, produce a tax summary and scenarios such as Roth conversions | Holistiplan | Annual plans priced by number of client households, 7-day free trial (Holistiplan) |
| Research and knowledge assistants | Answer questions from the firm’s research and policies | AI @ Morgan Stanley Assistant (in-house) | Not sold to outside firms |
| General assistants with finance connectors | Draft briefs, review portfolios, prepare meetings, check marketing language, connected to custodians and CRMs | Claude for Financial Advisors | Plugin free, usage billed on Anthropic’s plans (Radlof news) |
| Compliance and surveillance | Review emails and marketing for rule breaches, flag unusual activity | Features inside archiving and compliance platforms | Varies by vendor |
Meeting note-takers are the most visible category. Morgan Stanley’s Debrief, built on OpenAI models, takes notes in client meetings with the client’s consent, surfaces action items, drafts an email for the advisor to edit and send, and saves a note into Salesforce (Morgan Stanley). CNBC reported that clients must consent each time it is used and that one advisor in the pilot estimated it saved 30 minutes per meeting (CNBC). Independent firms buy the same kind of tool from vendors such as Jump and Zocks, which connect to advisor CRMs like Wealthbox, Redtail and Salesforce. One design difference matters for compliance: Zocks says it captures notes without keeping audio or video recordings (Zocks), while other tools store recordings that then have to be managed.
Tax return analysis is older than the chatbot wave. Holistiplan scans federal returns and transcripts and turns them into a tax report and scenario projections; its Premium tier adds state tax scenarios and multi-year Roth conversion projections (Holistiplan).
General assistants with finance connectors are the newest category. Anthropic’s Claude for Financial Advisors, launched September 14, 2026, bundles ready-made workflows for onboarding, meeting prep and follow-up, rebalance review, estate and tax briefs and compliance review, with connectors to Charles Schwab, Vanguard, BlackRock, Addepar, Orion, Wealthbox and others (Anthropic). Anthropic says investment recommendations, client communications and compliance determinations stay subject to human review and approval. Our news item on the launch has the details.
What the rules say
There is no separate US rulebook for AI in advice. Regulators apply the existing one.
FINRA says its rules are technology neutral and apply whether a firm builds its own AI tool or uses one embedded in a vendor’s product. Its 2024 notice names supervision (Rule 3110) and communications with the public (Rule 2210) as examples, and says the content standards for communications apply whether a human or a tool wrote them (FINRA Regulatory Notice 24-09). Its 2026 report adds recordkeeping and fair dealing, warns about hallucinations and bias, and suggests firms keep prompt and output logs, track which model version was used, and keep a human reviewing outputs (FINRA 2026 report).
The SEC lists AI in its fiscal 2026 examination priorities. Examiners will check whether what firms say about their AI is accurate, whether firms have policies to monitor and supervise AI used for tasks such as fraud detection, back-office work, anti-money laundering and trading, and whether automated advice matches each client’s profile, including for retail and older investors (SEC 2026 exam priorities). The SEC has already fined two advisers for overstating their AI: Delphia and Global Predictions paid $400,000 in total in March 2024 (SEC).
For an advisor choosing a tool, that translates into a short checklist:
- Consent. Get the client’s agreement before any tool listens to a meeting, and know your state’s call-recording rules.
- Review. Read every AI-written note, email and summary before it goes to a client or into the record. The advisor, not the vendor, is responsible for what is sent.
- Records. Make sure notes, drafts and, where kept, recordings are archived like other business records, and that deleting them does not break retention rules.
- Vendor due diligence. Ask where client data is stored, whether it is used to train models, what security certifications the vendor holds and how data is deleted at the end of the contract.
- Marketing. Do not describe your practice as “AI-driven” beyond what the tools really do; that is exactly what the SEC examines.
If you are a client
Most of these tools never touch your investment decisions, but they do touch your data: meeting conversations, tax returns, account details. It is reasonable to ask your advisor:
- Which AI tools do you use on my information, and did I agree to it?
- Is our meeting recorded, and if so, where is the recording kept and for how long?
- Does a person review AI-drafted recommendations and emails before I receive them?
- Is my data used to train the vendor’s models?
An advisor who can answer these plainly has probably thought about the rules above. If you use AI for your own money decisions rather than through an advisor, our guides to AI robo-advisors and the best AI personal finance assistants cover the consumer side.
How we checked this
We read FINRA’s and the SEC’s guidance and examination priorities, vendors’ own pricing pages and product announcements, and news coverage of Morgan Stanley’s rollout. Prices are US dollars before tax and change often, so confirm them with the vendor. Facts checked on September 25, 2026. This guide is general information, not legal or compliance advice; firms should confirm their obligations with their compliance team or counsel.
