What problem does it solve?
Insurance agents and brokers sell many products, each with product guides, underwriting rules and changing campaigns. Finding the right answer (is this condition acceptable, which rider fits, what documents are needed) means calling a helpdesk or searching intranets, which slows agents down, especially new ones.
At the same time, agents spend time on admin: preparing for meetings, writing follow ups, logging activity in the CRM and chasing leads that turn out to be poor. Insurers with large agency forces in Asia report sales or profit per active agent (Manulife and Prudential both disclose it), and at Zurich the client and broker facing commercial insurance teams mind more than 100,000 active opportunities in their CRM, so small time savings per person add up.
How does it work?
- Answer from approved content. Agents ask questions in their own words and get answers grounded in product guides, underwriting guidelines and procedures, with links to the source.
- Prepare the customer conversation. From CRM and policy data, the assistant suggests which customers to contact and why (a policy anniversary, a gap in cover, a life event) and drafts a personalized message for the agent to edit.
- Qualify leads. An AI agent contacts or screens incoming leads, validates contact details and interest, and passes qualified leads to the agent with notes.
- Handle the admin. After a meeting the assistant drafts notes and follow ups and updates the CRM for the agent to approve.
- Coach. Performance dashboards and call insights suggest where each agent can improve.
- Audience
- Employee facing
- Autonomy
- Assist
- Adoption
- Early adopters
- Channels
- Mobile app, Internal tools, Microsoft Teams
What is it worth?
Benchmarks are computed from the public deployments below: one data point per organization per KPI, with who made each claim.
| KPI | Median | Reported range | Data points | Claimed by |
|---|---|---|---|---|
| Users served | Not pooled | 300 | 1 | 1 vendor |
Value drivers: Employee productivity, Revenue growth, Customer experience, Compliance quality.
Indicative value
An insurer with 2,000 active tied agents
USD 1.1 million to USD 7.4 million
Value of agent time redeployed to selling per year
How this is calculated
Formula: agents * hoursSavedPerWeek * weeksPerYear * valuePerHour. The low scenario uses every low input, the high scenario every high input.
| Input | Low | High | Basis |
|---|---|---|---|
| Active agents using the assistant agents, agents | 2,000 | 2,000 | The reference insurer. |
| Hours saved per agent per week hoursSavedPerWeek, hours per week | 0.5 | 2 | Editorial assumption for product lookups, meeting preparation and follow up drafting. Replace with your own time study. |
| Working weeks per year weeksPerYear, weeks | 46 | 46 | Editorial assumption. |
| Value of an agent hour redeployed to selling valuePerHour, USD per hour | 25 | 40 | Editorial assumption, replace with your own figure for agent earnings per hour. |
What it leaves out: Time value only; it assumes saved time goes into customer contact. It leaves out any sales uplift, licence and platform costs, content maintenance and the compliance review of AI drafted customer messages.
Who already uses it?
5 public deployments, strongest evidence first. Grades: A regulator or audit, B the organization itself, C vendor case study, D anonymous or estimate.
Manulife
Canada · Insurance · 2025
Manulife's annual reports describe a generative AI sales tool in Singapore and Japan that drafts personalized engagement strategies for agents from customer needs, demographics and transaction history. By 2025 it had deployed GenAI sales enablement across nine markets and all four operating segments, with engagement insights, automated email drafting and real time coaching; in Hong Kong it launched AI Sales Pro for agents, and in Indonesia, Singapore and Japan AI assistants give agents faster access to product and policy information. In U.S. Retirement (part of its Global Wealth and Asset Management segment), Manulife says an AI sales enablement solution reduced time spent on information searches and tripled the number of sales opportunities compared with 2024.
No outcome disclosed.
Sun Life
Canada · Insurance · 2025
Sun Life's 2025 annual report says AI tools cut median response times for individual insurance applications in a target segment by close to half, including a 50% increase in straight through underwriting. In its Client Contact Centre, generative AI raised chatbot containment by 16 percentage points year over year, and advisors use a generative AI Notes Assistant Tool; in Hong Kong it launched Advisory Buddy, a GenAI chatbot inside its Advisor Workbench. In Malaysia almost two thirds of clients received automated underwriting decisions within two hours.
- Cycle time reduction: about 50%, median response time for individual insurance applications, target segment, 2025
"Leveraged AI tools to do more for our Clients including reduced median response times for individual insurance applications for a target segment by close to half (including increasing straight-through underwriting by 50%), increased chatbot containment rate (up 16 percentage points year-over-year) with generative AI tools in the Client Contact Centre, and enhanced advisor productivity using a generative AI-powered Notes Assistant Tool."
Claimed by: organization
Waterdrop
China · Insurance · 2025
Waterdrop, a Chinese online insurance distribution and health services platform, runs a suite of AI applications called Waterdrop Guardian that either talk to users directly or support its online insurance consultants. In its second quarter 2025 results it reported that its AI Customer Service Agent resolved 60% of inquiries on first contact, that its Life Planner Copilot handled 300,000 product consultations for consultants, and that premiums facilitated by its AI Medical Insurance Expert rose 155% from the previous quarter (the release does not say whether this tool talks to users directly or supports consultants). It also launched KEYI.AI, a real time AI underwriting assistant for consultants. Its 2023 annual report describes an earlier LLM powered AI Insurance Consultant that was tested internally in medical insurance scenarios, an internal test rather than a sales deployment.
- First contact resolution: 60%, second quarter 2025, AI Customer Service Agent
"‘AI Customer Service Agent’ resolved 60% of inquiries on first contact, enhancing user experience."
Claimed by: organization
Prudential plc
Hong Kong SAR China · Insurance · 2024
Prudential uses AI across its Asian agency force. An AI talkbot validates and enriches leads before they reach agents in Singapore and the Philippines; for the Philippines, launched in the second half of 2024, Prudential reported an initial result of 98 per cent of the talkbot's validated leads being adopted by agents for follow up. In 2025 it launched PruAction in Singapore, a generative AI performance management platform with real time insights for agents, and introduced a Health AI chatbot in Singapore that helps agents access information more quickly. Prudential lists these tools among its agent productivity initiatives but does not attribute its productivity figures to them.
No outcome disclosed.
Zurich Insurance Group
Switzerland · Insurance · 2025
Zurich's commercial insurance teams manage more than 100,000 active opportunities in Dynamics 365, and switching applications to copy updates from email into the CRM left data at risk of going stale. With Microsoft 365 Copilot for Sales, 300 users create and update contacts and link emails to opportunities from Outlook, get summaries of relationships and long email threads, and draft emails. Zurich estimates about 14,000 hours saved over the next year; that is an estimate, not a measured result. The story also says user feedback indicates the tool improves Zurich's sales and retention ratios, without figures.
- Users served: 300
"Copilot for Sales has quickly become a critical productivity tool for Zurich’s 300 Copilot users, who find even more benefits as they continue to work with it in Outlook."
Claimed by: vendor
How do you implement it?
A model agnostic playbook: what to prepare, the order to build in, and what goes wrong.
Data you need
- Current product guides, underwriting guidelines and sales procedures with owners
- CRM and policy data per agent's book, with contact consent flags
- Approved message templates and compliance rules for customer communications
- Lead sources and qualification criteria
Systems to integrate
- Agent portal or mobile app
- CRM (for example Dynamics 365 or Salesforce)
- Policy administration for in force data
- Email, calendar and Microsoft Teams
- Telephony or messaging for lead qualification
Complexity: Medium
Knowledge answering on approved content is well understood. Personalized engagement and lead qualification need clean CRM and policy data, consent for contact, and compliance review of what the assistant suggests agents say.
- 1
Start with the question agents ask most
Mine helpdesk tickets and calls from agents to find the top product and underwriting questions and build the first version around them.
- 2
Own the content
Give every product guide and guideline an owner and a review date; stale content is the main reason agents stop trusting the assistant.
- 3
Keep agents in control of customer messages
Drafts are suggestions the agent edits and sends; nothing goes to a customer automatically, and compliance approves the templates behind them.
- 4
Pilot in one market and measure productivity
Compare active agents using the assistant with a similar group on activity, conversion and time to first sale for new agents.
- 5
Add lead qualification carefully
When an AI contacts leads directly it becomes customer facing: add disclosure, consent checks and a clean handover to the agent.
Guardrails
- Answers only from approved, current content, with sources shown
- No automatic sending of customer messages; agents approve every draft
- Customer data access limited to the agent's own book
- Contact suggestions respect marketing consent and do not target vulnerable customers inappropriately
- Clear disclosure when an AI agent contacts a lead directly
KPIs to instrument
- Weekly active agents as a share of the agency force
- Questions answered without escalation to the helpdesk
- Share of qualified leads followed up by agents
- New business per active agent, assistant users versus comparison group
- Compliance findings on AI assisted communications
Human in the loop
Agents decide what to recommend and send. Distribution compliance approves templates and reviews a sample of AI assisted communications, and product owners keep the content current.
Common failure modes
- Confident answers from old guides
- An agent quotes a withdrawn product feature to a customer. Retire content on schedule and show the document date with every answer.
- Personalization that crosses a line
- Suggestions use data the customer did not expect to be used for sales. Check purpose and consent for every data source.
- Adoption stalls after launch
- Agents try it once and go back to the helpdesk. Measure weekly active use and fix the top unanswered questions every week.
- Productivity claims without a comparison
- Agent productivity rises for many reasons. Use a comparison group before attributing gains to the assistant.
What are the risks and rules?
EU AI Act
Depends on design
An employee facing assistant for knowledge answers and drafting is not listed in Annex III and is minimal risk. A lead qualification agent that talks to customers must tell them they are dealing with AI (Article 50). Using performance insights to monitor and evaluate individual agents, or to allocate leads based on their behaviour or traits, is high risk under Annex III point 4(b), and any component that does risk assessment or pricing of life or health insurance for individuals is high risk under Annex III point 5(c).
Guidance
- Insurance Distribution Directive (IDD) (European Insurance and Occupational Pensions Authority, Europe). Conduct, product information and demands and needs rules apply to what agents tell customers, including when an AI drafted it.
- Opinion on Artificial Intelligence governance and risk management (European Insurance and Occupational Pensions Authority, Europe). Addressed to national supervisors; explains how existing insurance rules on governance, risk management and fair treatment of customers apply to AI systems that are not high risk under the AI Act.
- Annex III, high risk AI systems referred to in Article 6(2) (European Union, Europe). Point 4(b) covers monitoring and evaluating the performance of people in work relationships; point 5(c) covers risk assessment and pricing in life and health insurance.
Controls to put in place
- Content ownership and review dates for every source document
- Compliance approval of message templates and a monthly sample review
- Access control by agent book and role
- Consent checks before contact suggestions
- Usage and outcome monitoring by market
- Performance insights coach agents and are not used on their own for decisions on contracts, commission or lead allocation
Frequently asked questions
- What do insurers use AI assistants for in their agency force?
- Manulife has deployed GenAI sales enablement across nine markets, with engagement insights, email drafting and coaching; Prudential uses an AI talkbot to validate leads and a GenAI performance platform for agents; Sun Life gives advisors a GenAI chatbot and a notes assistant.
- Is there evidence it makes agents more productive?
- Some. Waterdrop reports its Life Planner Copilot handled 300,000 product consultations for its online consultants, and Prudential reports an initial result of 98 per cent of leads validated by its talkbot in the Philippines being adopted by agents for follow up. None of the insurers cited here publish productivity figures against a comparison group.
- Who is responsible for what the agent tells the customer?
- The agent and the insurer or broker, as today. AI drafts do not change distribution rules under the IDD or conduct rules such as the FCA Consumer Duty, so templates need compliance approval and agents should edit before sending.
How to cite this page
Blits.ai AI Use Case Library, "AI assistant for insurance brokers and agents", last verified 27 September 2026, https://www.blits.ai/ai-use-cases/insurance-broker-and-agent-assistant. Licensed under CC BY 4.0. Method: how we verify use cases.
Changelog
- 27 September 2026: First published