# How should an independent insurance agency start using AI? > Pick one approved tool, set data rules, and start with policy review recaps and claim-status replies that never promise what the carrier decides. Source: https://learn.ynteractive.com/content/questions/how-insurance-agencies-start-with-ai · Updated 2026-10-22 · Free from Stepthrough (https://learn.ynteractive.com) By [Robert Breen](https://learn.ynteractive.com/content/about) · Published October 22, 2026 · Updated October 22, 2026 · 4 min read An independent agency should start the way it would with any new system: the owner picks one approved AI tool on a business plan, writes down what client data may and may not go in, and then rolls it out on two jobs where the time savings are clear and the risk is controlled, policy review recaps and claim-status replies. [Turn Every Policy Review into Clear Next Steps](https://learn.ynteractive.com/content/chatgpt-policy-review-meeting-recap-insurance-agency) shows the first one, including a separate list of questions only the carrier can answer. The owner's view matters because the risk in an agency is not the AI. It is a producer or CSR on a personal account pasting a client's policy details, or a reassuring email that says "you're covered" before the carrier has confirmed it. ## What should the agency owner decide first? 1. **One tool, one plan.** A ChatGPT or [Claude](https://learn.ynteractive.com/content/glossary/claude) business plan, or the AI your agency management system or Microsoft 365 already includes. Personal accounts are out for client work. 2. **Data rules.** Policy numbers, dates of birth, driver's license numbers, bank details and claim medical information stay out unless your setup is approved for them. 3. **The "carrier decides" rule.** No AI-drafted message may confirm coverage, payment, rates or claim outcomes. Those come from the carrier, and the email should say "we're checking" until they do. 4. **Review.** A licensed person reads every client-facing message before it goes. 5. **Fit it into your security program.** Many states have adopted insurance data security laws based on an NAIC model that require licensees, including agents, to maintain an information security program. Add your AI tool to it like any other vendor. ## Which two jobs should we start with? **Annual policy review recaps.** A half-hour review with a small-business client covers new equipment, new hires, new exposures and questions about coverage. By the next day, half of it is forgotten, [endorsements](https://learn.ynteractive.com/content/glossary/policy-endorsement) go in late and quote requests never get made. AI turns the transcript into decisions, an [action items](https://learn.ynteractive.com/content/glossary/action-items) table with owners and dates, and a list of carrier questions. **Claim-status replies.** Clients with an open claim want a payment date, a rental and a promise about their rate. The agency cannot decide any of those. AI drafts a kind, clear reply from the facts you give it, and you check every promise against those facts. Both jobs are frequent, follow a pattern and are easy to check. They are also exactly where an unchecked "you're covered" turns into an errors and omissions problem, so a good routine protects the agency as well as saving time. ## What does this look like in a real lesson? In [Turn Every Policy Review into Clear Next Steps](https://learn.ynteractive.com/content/chatgpt-policy-review-meeting-recap-insurance-agency), Kestrel Point Insurance Agency has just finished an annual review with Fenwick Print & Sign. You paste the Google Meet transcript into ChatGPT under a request for five things: a summary, the decisions, an action items table with owner and due date, open questions, and questions for the carrier. Two dates come back as **Not stated**. You tell ChatGPT that Tasha will send the payroll estimate by Wednesday and Dev will have the cyber quotes by October 23. Then you ask for a recap email to the client that promises no coverage the carrier has not confirmed. In [Reply Faster: Claim-Status Emails](https://learn.ynteractive.com/content/chatgpt-reply-claim-status-emails-insurance-agency), Denise at Stonebridge Insurance Agency hit a deer a week ago and asks when she will be paid, whether a rental is covered and whether her rate will go up. The facts you give ChatGPT include who decides payment, what her declarations page shows for a rental and your agency's rule. The promise check shows six of eight claims match a fact, the rental line is only partly supported and the rate promise is not supported at all. You use the suggested wording and send an honest reply from Gmail. Before either goes live, staff should know the habits in [Use AI at Work Without Leaking Client Data](https://learn.ynteractive.com/content/safe-ai-at-work-what-not-to-paste): training off, placeholders for personal details, and a check for made-up promises. ## How do we roll it out across the agency? - **Week 1:** owner and one CSR use it on recaps and claim replies, and save the prompts that work. - **Week 2:** share the saved prompts with the team and the one-page rules. - **Week 3 onward:** add renewal reminders, certificate request replies and new-client welcome emails, one at a time. The [insurance agent start guide](https://learn.ynteractive.com/content/start/insurance-agent) and [5 tested ChatGPT uses for insurance agencies](https://learn.ynteractive.com/content/blog/chatgpt-for-insurance-agencies) list more tasks to add next. ## What mistakes should agency owners avoid? - **Letting staff choose their own tools.** You cannot protect data you cannot see. - **Recaps that say "you're covered."** Keep carrier questions separate. - **Pasting full applications or loss runs** into an unapproved tool. - **Skipping review because the draft sounds right.** That is when promises slip in. - **Starting with a chatbot on the website.** Learn on internal work first. ## Key takeaways - The owner sets one tool, data rules and a "carrier decides" rule first. - Start with review recaps and claim-status replies. - Keep carrier-only questions in their own list, and say "we're checking" until confirmed. - A licensed person reviews every client-facing message. ## Frequently asked questions **Can AI quote insurance or compare carriers?** AI can organize what carriers sent and draft a client-friendly comparison. Coverage advice and the recommendation stay with a licensed agent, and quotes come from the carriers' own systems. **Does AI create E&O risk for agencies?** Unchecked messages can, especially ones that confirm coverage or predict claim outcomes. Clear rules, a carrier-questions list and human review lower that risk. Ask your E&O carrier whether it has guidance. **Do we need a special insurance AI tool to start?** No. A general assistant on a business plan handles recaps and replies well. Check what your agency management system already includes before you buy anything new. ## Learn it hands-on Free, interactive lessons: you do every step yourself in a practice copy of the real tool, it checks your work as you go, and nothing touches your real accounts. - [Turn Every Policy Review into Clear Next Steps using ChatGPT](https://learn.ynteractive.com/content/chatgpt-policy-review-meeting-recap-insurance-agency): Turn the transcript of a small-business client's annual policy review into the decisions, an action items table, the questions only the carrier can… (about 7 minutes) - [Reply Faster to Claim-Status Emails: ChatGPT for Independent Insurance Agencies](https://learn.ynteractive.com/content/chatgpt-reply-claim-status-emails-insurance-agency): A client with an open auto claim wants a payment date, a rental and a promise about her rate. Give ChatGPT the thread and your agency's facts, catch… (about 10 minutes) - [Use AI at Work Without Leaking Client Data using ChatGPT](https://learn.ynteractive.com/content/safe-ai-at-work-what-not-to-paste): Turn off model training, strip a tenant's bank details and address out of a prompt, then catch the two promises ChatGPT made up before the reply… (about 7 minutes) [Start the free lesson](https://learn.ynteractive.com/modules/insurance-review-meeting-summary) ## AI tools in this answer Each one in the Stepthrough AI tools directory: what it does, who it is for and the free lesson where there is one. [ChatGPT](https://learn.ynteractive.com/content/ai-tools/tool/chatgpt) · [Claude](https://learn.ynteractive.com/content/ai-tools/tool/claude) [The full AI tools directory](https://learn.ynteractive.com/content/ai-tools) ## Related terms Plain-English definitions, each with an example from a free lesson. [Claude](https://learn.ynteractive.com/content/glossary/claude) · [Policy endorsement](https://learn.ynteractive.com/content/glossary/policy-endorsement) · [Action items](https://learn.ynteractive.com/content/glossary/action-items) · [Actual cash value (ACV)](https://learn.ynteractive.com/content/glossary/actual-cash-value) · [AI note-taker](https://learn.ynteractive.com/content/glossary/ai-note-taker) · [AI policy](https://learn.ynteractive.com/content/glossary/ai-policy) [The full AI glossary](https://learn.ynteractive.com/content/glossary) ## Related prompts - [Action items table prompt](https://learn.ynteractive.com/content/prompts/action-items-table-prompt): Paste a meeting transcript or rough notes and get a table of every task, owner and due date, with "Not stated" wherever nobody… - [Annual coverage review invite prompt](https://learn.ynteractive.com/content/prompts/annual-coverage-review-invite-prompt): Invite an insurance client to a short annual review: why now, what to have handy, two times to pick from. No scare tactics and… - [Certificate of insurance request reply prompt](https://learn.ynteractive.com/content/prompts/certificate-of-insurance-request-reply-prompt): A ChatGPT prompt to answer a certificate of insurance request: the details you need from the client, what you will send and… [All free prompt templates](https://learn.ynteractive.com/content/prompts) ## Related questions - [Can lawyers use ChatGPT without breaking their ethics rules?](https://learn.ynteractive.com/content/questions/can-lawyers-use-chatgpt-ethically): Yes, if you understand the tool, protect client confidences, supervise staff, verify every citation and bill fairly. Not legal… - [How do consultants use AI on client work without exposing client data?](https://learn.ynteractive.com/content/questions/how-consultants-use-ai-client-work): Consultants use AI for kickoff recaps, market scans and decks, on business plans that do not train on data, with client names… - [Where should a paralegal start with AI if the firm has no AI policy yet?](https://learn.ynteractive.com/content/questions/how-paralegals-start-with-ai): Ask a supervising attorney first, keep client confidences out, and practice on non-confidential tasks until the firm sets rules… - [How should a finance team start using AI without breaking controls?](https://learn.ynteractive.com/content/questions/how-to-start-using-ai-in-finance): Treat AI as a preparer, never an approver: written rules, an approved tool, a person who reviews and evidence you keep. A plan… - [Where should an HR team start with AI if nobody on the team has used it yet?](https://learn.ynteractive.com/content/questions/how-to-start-using-ai-in-hr): Start with three team rules on employee data, then one low-risk drafting task, then a policy answer bot. A plain order for HR… - [How should L&D roll out AI training to the whole company?](https://learn.ynteractive.com/content/questions/how-to-train-staff-to-use-ai): Start with a one-page AI policy, then hands-on practice on real tasks, then role-based tracks. An L&D rollout plan with free… [More questions: AI by department](https://learn.ynteractive.com/content/questions/by-department) · [All AI questions](https://learn.ynteractive.com/content/questions) · [All free tutorials](https://learn.ynteractive.com/content)