AI for insurance agents: where it helps, and what stays with you
Renewals don't get missed because nobody cared. They get missed because the detail the client mentioned in March is in a thread nobody reopened. This guide maps the AI tools agents run into, where they save time, what stays with the licensed agent, the rules worth knowing, and a simple way to keep what each client told you.
Short answer: AI is useful to an insurance agent for the writing and reading around the work: first drafts of renewal letters and follow-ups, plain-English explanations, summaries of long material you are allowed to share. It is not a substitute for reading the policy yourself, for the coverage recommendation, or for the conversation at renewal. And it knows nothing about what a client told you on the phone in March unless someone wrote it down.
That last gap is the one that costs accounts. The client mentioned the teenager getting a licence, the new roof, the side business run from the garage. It came up on a call, or in the third email of a thread about something else. Nine months later the renewal arrives, and the detail that should have shaped it is in a place nobody reopened.
One disclosure before going further. This guide is published by Equerry, a personal assistant on your iPhone. Equerry is not a quoting, rating or policy-servicing tool, and it never calls, texts or messages your clients. Much of what comes up when you search this topic is AI that phones or texts people on an agency's behalf; Equerry is not that. Nothing on this page is legal, compliance or coverage advice. The first part covers the landscape on its own terms; Equerry comes in near the end, where it fits.
The four kinds of AI an agent runs into
| Kind | What it is for | What to ask first |
|---|---|---|
| AI phone and messaging agents | Answering inbound calls, booking appointments, placing or replying to calls and texts for the agency | Consent for automated calls and texts, what it says in your name, where the call recordings and transcripts go |
| General chat assistants (ChatGPT, Claude, Gemini) | First drafts, rewording, explaining a process, summarizing material you may share | Your agency's policy, which plan you are on, the training and retention settings |
| AI inside agency software | Features added to agency management systems, CRMs and carrier portals; they vary by vendor | What the vendor's documentation says it does with client data |
| Client-record support | Keeping what each client said and what is coming up, and being briefed before the next conversation | What it can reach, how you add to it, whether you can see and delete what it kept |
The first kind fills most of the search results for this topic. Vendors describe it in their own words: Sonant's site calls its product "The AI Receptionist For Insurance" and promises to "Turn routine incoming calls into revenue in less than 30 days" (read 25 September 2026). It is named here as an example of the category, not a recommendation. If your agency is weighing one, the rules section below has the FCC ruling on AI-generated voices in calls.
Where AI genuinely saves an agent time
These are the jobs where a general assistant earns its place, with the check that goes with each.
- Renewal and follow-up letters. A warm, specific note after an annual review, or a renewal letter that explains a premium change in plain words. Give it the facts; you check every figure against the carrier's documents before it goes out.
- Explaining insurance in plain English. What a deductible is, how an umbrella policy sits on top of auto and home, why a claim takes the time it does. A draft explanation you edit is faster than a blank page, as long as it describes insurance in general and not what a particular policy covers.
- Summaries of long material, as a map. A long policy form or endorsement can be summarized into the sections a client will ask about. Use the summary to find where to look; the policy wording, not the summary, is what you advise from.
- Prospecting and content. A newsletter paragraph before hurricane season, a post about reviewing coverage after a home renovation, a script for a first call. General content with no client details in it is the easiest place to start.
- Getting ready for a review. Turning your own notes on a client into a short list of questions to ask. The notes have to exist first, which is the subject of the second half of this page.
What stays with the licensed agent
The pattern: AI drafts, you answer for it. A few areas deserve more than a proofread.
- What the policy covers. Summaries miss exclusions, endorsements and definitions. Coverage questions are answered from the policy and the carrier, not from a chat window.
- Quotes and rates. Premiums come from the carrier's rating, not from a general assistant, which can produce a confident number that means nothing.
- The recommendation. Which coverage, which limits, whether a product suits this client: that is the licensed judgment the client is relying on, and your state's rules and your carriers set how it has to be made and recorded.
- What goes out in your name. A letter, an email or a message drafted by a tool is still yours once you send it. Your agency and carriers may have rules on what client communications can say; reread anything that describes coverage.
- The relationship. The claim call after a house fire, the conversation about a lapse, the client who wants to cancel. Nobody wants those handled by a draft.
The rules worth knowing, as information
Insurance is regulated state by state, and your own state's insurance department, your agency and your carriers decide what applies to you. What follows is a short, dated reading list from primary sources, not legal or compliance advice.
- The NAIC AI model bulletin is addressed to insurers. The National Association of Insurance Commissioners adopted its model bulletin, "Use of Artificial Intelligence Systems by Insurers", on 4 December 2023. It is addressed to "All Insurers Licensed to Do Business In (Insert Name of Jurisdiction)", and its stated purpose is to remind insurers "that decisions or actions impacting consumers that are made or supported by advanced analytical and computational technologies, including Artificial Intelligence (AI) Systems (as defined below), must comply with all applicable insurance laws and regulations." The NAIC calls it "a guiding document" rather than a model law or regulation, and it is written for a state insurance department to issue under its own name, so check whether your state's department has issued it.
- Existing law still applies. The NAIC's March 2026 issue brief on AI and state insurance regulation, which is about insurers' own use of AI, says: "AI is a tool used in underwriting, pricing, claims, fraud detection, and utilization management. It does not alter insurers’ legal obligations. Existing state insurance laws apply regardless of whether decisions are made by humans, algorithms, or third-party vendors."
- Client information. The FTC's Gramm-Leach-Bliley Act page (read 25 September 2026): "The Gramm-Leach-Bliley Act requires financial institutions – companies that offer consumers financial products or services like loans, financial or investment advice, or insurance – to explain their information-sharing practices to their customers and to safeguard sensitive data." How that applies to your agency and the tools you use is a question for your agency and your state, not for this page.
- Automated calls and texts. If your agency is considering an AI phone agent, the Federal Communications Commission's Declaratory Ruling FCC 24-17 (adopted 2 February 2024, released 8 February 2024) says: "In this Declaratory Ruling, we confirm that the TCPA’s restrictions on the use of “artificial or prerecorded voice” encompass current AI technologies that generate human voices." Consent requirements for automated calls and texts are a question for your agency's counsel before any tool dials a client.
Before you paste client details into a chatbot
- Ask your agency first. Some agencies and carriers have a policy on which tools may hold client information. A personal account on a consumer plan may not be the approved answer. Carriers you are appointed with may have their own rules on AI tools used with their business; ask.
- Share what the job needs. A plain-English explanation of an umbrella policy needs no client name, address, date of birth or claims history.
- Check the tool's settings. On personal ChatGPT plans, a setting under Settings, then Data controls, decides whether new conversations can be used to train OpenAI's models, and OpenAI treats its business workspaces differently by default. Our guide for lawyers walks through that setting step by step; the same questions apply to every other assistant.
- Remember what the setting doesn't answer. Training is one question. Who else can access the information, how long it is kept and whether your agency allows it at all are the others.
The renewal problem: keeping what each client told you
Every agent carries a running file in their head: who is adding a driver in the spring, who mentioned a boat, who asked you to look at their umbrella limit "next time". It is the least glamorous part of the book and the part renewals depend on. You can keep it well without any AI, and most good agents do some version of this:
- Write the note in the first five minutes after the call. Three lines: what changed in their life, what they asked for, what you owe them and by when. Our guide to keeping a record of client calls covers the habit in detail.
- Put the note where the renewal will find it. In the account in your agency management system, the same day. The detail mentioned in March only helps if it is in front of you in December.
- Work renewals from a list, early. A list of renewals coming up in the next 60 to 90 days, reviewed each week, gives you time to act on what clients told you rather than discovering it on the renewal date.
- Read the file before the call. Five minutes with the last note and the last few emails before a review. Clients notice when you remember.
The weak point is always the email. A client mentions the new roof in the fourth reply of a thread about a billing question, and the note in the system never gets written.
Where Equerry fits: a personal assistant for the client record
Equerry is a personal assistant who helps you get things done, and for an agent that means the fourth kind of tool on the map. It does not quote, rate, bind, advise or contact anyone. It keeps track of what you send it, and it brings things back when you need them, the way a good assistant keeps the file straight.
- Send it the policy PDFs. Share a declarations page PDF or a screenshot from your iPhone's share sheet, or type a note. It's there when you ask about that client later.
- CC it on the client thread. Your Equerry has its own email address at mail.getequerry.app. Add it to the CC line and, for as long as it stays on the thread, later replies keep arriving; or forward just the one email you need. It never asks for your inbox, and the original stays in your own mailbox.
- Debrief the call by voice. After a review call, dictate a voice note in the Equerry app: the new driver, the roof, the question about umbrella limits. It is transcribed so you can ask about it later. Equerry does not join or record your calls; the voice note is yours, made after the call.
- Put the renewal on a date. Ask Equerry to remind you on a date you pick, such as the date sixty days before a client's renewal, and a Reminder brings it back on the day. A Watcher can check what you've sent on a schedule you pick (hourly, every six hours or daily) for something new on a topic, such as a carrier's renewal offer for one account, and ping you when it finds it.
- Ask in plain words. "When does the Alvarez homeowners policy renew, and what did she ask for?" is answered from the notes, emails and voice notes you gave it. When an answer draws on a note you saved, it carries a source chip: tap it to see the passage Equerry used and open the original.
- Walk into the review briefed. If you allow calendar access, Equerry keeps a snapshot of your events from today through the next 14 days (titles, start and end times, locations and attendee display names, never event notes, meeting links or email addresses), and a Morning Brief, if you set one up, pulls together what you've saved that bears on the day's appointments.
- See what it kept. Settings → Your memory lists the facts it has kept. For an email you sent it, what Equerry remembered from it is listed there, each line marked as from that email, and you can delete any line.
| Job | AI phone or messaging agent | General chat assistant | Equerry |
|---|---|---|---|
| Answering the agency's phone | What it's built for | Not built for it | No, and it never contacts clients |
| First draft of a renewal letter | Not what it's built for | Good at it, from facts you give it | Not what it's for |
| What a client mentioned in March | Depends on what the vendor records and where it goes | Only what you typed, uploaded or connected to it | A voice note, typed note or CC'd thread, asked about later |
| Remembering the renewal date | Varies by vendor | Depends on the plan and what you set up | A Reminder on the date you choose |
Honest limits. Equerry doesn't connect to your agency management system, your CRM, a rater or any carrier portal, doesn't join or record calls, and never calls, texts, messages or emails a client on your behalf. It knows what you send it, plus your calendar if you allow it, and nothing else; it cannot find the email you never forwarded. Anyone on a thread can see the Equerry address on the CC line, so forward instead where that would be awkward. And sending something is still sharing it: what you send is processed to make it searchable and answerable, including by the outside providers named on our AI disclosure page. Equerry makes no claim about meeting any insurance, privacy or telemarketing rule. Whether client material goes into any tool, this one included, is a decision for you under your agency's policy, your carriers' requirements and your state's rules. What it changes is scope: the question is about the specific threads and notes you chose to send, rather than a whole mailbox.
If most of your client work happens over email, read what an AI email assistant can and can't do next. Advisors who also sell insurance may want AI for financial advisors, which covers the FINRA and SEC side.
Sources, all read 2026-09-25. NAIC, NAIC Model Bulletin: Use of Artificial Intelligence Systems by Insurers (adopted 4 December 2023); NAIC Members Approve Model Bulletin on Use of AI by Insurers (4 December 2023); Artificial Intelligence and State Insurance Regulation (issue brief, March 2026). Federal Trade Commission, Gramm-Leach-Bliley Act. Federal Communications Commission, Declaratory Ruling FCC 24-17 (released 8 February 2024). OpenAI Help Center, Data controls in ChatGPT. Sonant, sonant.ai, named as a category example only; not a ranking or an endorsement. Vendors change their products and pages often; where a vendor's current page disagrees with this one, the vendor's page is right. This page is general information, not legal, compliance or coverage advice.
Frequently asked questions
How can insurance agents use AI?
Mostly for the writing and reading around the sale, not the advice itself. A general chat assistant is a good first-draft writer for renewal letters, follow-up emails after a review and plain-English explanations of a process, and a fair summarizer of long material you are allowed to share with it. What it cannot do is decide what a client should buy, or tell you what the policy actually covers: that reading, and the recommendation, stay with the licensed agent. The other job worth solving is the client record, so that what a client mentioned in March is in front of you at renewal.
Can insurance agents use ChatGPT?
Whether you may is your agency's call first, and some carriers and agencies set their own rules for client information, so ask before you paste anything about a client into it. For work with no client details in it (a newsletter paragraph, a script for explaining deductibles, a checklist for a new hire) many agents use it the way everyone else does. On a personal ChatGPT plan, check the training setting under Settings, then Data controls; our guide for lawyers walks through it step by step. Treat every output as a draft you answer for.
What AI tools do insurance agencies use?
Roughly four kinds, named here as categories rather than a ranking. AI phone and messaging agents that answer or place calls and texts for an agency (Sonant, for example, an inbound call answering service, describes itself as "The AI Receptionist For Insurance" on its own site, read 25 September 2026). General chat assistants for drafting and summarizing. AI features built into agency management systems and CRMs, which vary by vendor, so ask yours. And tools that keep the client record, so you can ask what a client said and when their policy renews. Which ones an agency may use depends on its carriers, its policies and its state.
Is it OK to put client policy details into an AI tool?
Check before you do. Policy details, claims history and personal information are what the privacy rules around insurance are about: the FTC describes the Gramm-Leach-Bliley Act as requiring financial institutions, including companies that offer insurance, to explain their information-sharing practices and to safeguard sensitive data. How that applies to you is a question for your agency, your carriers and your state, not for a vendor's marketing page, and this is not legal advice. Practically: share only what the task needs, prefer a tool your agency has approved, and know what each tool does with what you send it.
Will AI replace insurance agents?
No one can promise either way, and the NAIC material quoted on this page is about how insurers use AI, not about replacing agents. What is changing is the mix of work: some phone answering, first drafts and summaries are getting faster, while the coverage conversation, the judgment about what fits a client and the relationship at renewal stay with people. How that plays out in a given agency is a business decision.
Last updated 2026-09-25.
All product names, logos, and brands are property of their respective owners. Equerry is not affiliated with, endorsed by, or sponsored by any other product mentioned; comparisons are provided for informational purposes.