Wise Agents
  1. Home
  2. Industries
  3. Insurance
Industry · Insurance

AI agents for insurance companies

Claims, policy questions and complaints are where insurers drown in documents. We build agents that prepare that work for your handlers. Risk assessment and pricing for life and health insurance is high-risk under the AI Act, so our agents stay off that line unless you build to it with us.

Where the work piles up

The problems we hear first

Claims that arrive incomplete

Photos without receipts, forms without policy numbers: handlers spend the first day of a claim asking for what is missing.

Coverage questions

Customers and brokers ask what is covered. The answer is in the policy wording, and it has to be exactly the wording.

Complaints against a clock

Each complaint needs a registered entry, an investigation and a plain-language reply within your policy’s deadline.

One standard in several languages

Cross-border books need the same quality of answer in every language, and the same record behind it.

The agents we build

Each one prepares. A person decides.

Examples of what we build in this sector, not client work. For each one: where a person approves, and what it never does.

Agent we build

Claims intake assistant

Reads the claim and attachments, checks cover against the policy wording, asks for what is missing and drafts a payout proposal with its sources.

A person approves
A claims handler approves every payment or refusal.
It never
Pay, refuse or reduce a claim.
Agent we build

Coverage answers

Answers customers and brokers from the approved wording only, names the clause it relied on, and hands over when a question needs advice.

A person approves
Advice and exceptions go to your licensed staff.
It never
Recommend a product. Demands and needs stay with your staff (IDD Art. 20).
Agent we build

Complaints desk

Registers the complaint, classifies it, gathers the file and drafts a reply, with the deadline on every draft.

A person approves
A complaints handler decides and sends every reply.
It never
Decide a complaint or send a reply.
Agent we build

Underwriting file preparation (non-life)

Collects the submission, checks it against your underwriting guidelines and lists the gaps.

A person approves
The underwriter decides the risk and the price.
It never
Price or accept a risk, or assess a person for life or health cover.
A use case, as a run

One job, step by step, with the person inside the flow.

Sample data · scripted

Complaint about a claim decision

A policyholder complains that a water-damage payout was reduced. The agent prepares the file and a draft reply; a complaints handler decides.

Sector
Insurance
Risk class
Not listed in Annex III; reasons kept
Who decides
A complaints handler
  1. Complaint receivedBy email, in German: the payout was reduced and the policyholder doesn’t understand why (sample).
    Done
  2. Registered and classifiedCategory: claim settlement. The deadline from your complaints policy is shown on the file.
    Done
  3. Claim file gatheredThe claim, the handler’s decision note and the policy section it relied on; nothing else.GDPR Art. 5(1)(c)
    Done
  4. One gap foundThe decision note cites an exclusion that the letter to the customer never named.
    Flagged
  5. Reply draftedIn plain German: the clause, why it applied, and the escalation route your policy names.
    Drafted
  6. Waiting for a complaints handlerThe handler upholds, partly upholds or rejects the complaint, and edits the reply.UpholdPartly upholdRejectAI Act Art. 14GDPR Art. 22
    Waiting for a person
  7. Run recordedThe file, the draft, the handler’s changes and the date it was answered.AI Act Art. 12
    On record

A sample, scripted on invented data: no model is called and nothing here is client work.

The rules that apply

Where the law draws the line for you

Each rule with the day we read it. An engineer’s reading for your counsel to confirm, not legal advice. How we read the AI Act.

RuleWhat it says (our summary; quotes are verbatim)What it means for an agent
AI Act Annex III 5(c)Read 12 November 2026Lists AI for “risk assessment and pricing in relation to natural persons in the case of life and health insurance”.An agent that assesses or prices a person for life or health cover is high-risk from 2 December 2027. Claims intake, property and motor are not in this point.
AI Act Art. 27Read 12 November 2026Deployers of Annex III 5(b) and 5(c) systems carry out a fundamental rights impact assessment before first use.If you build a 5(c) system, the assessment comes before go-live; we prepare its inputs.
IDD Art. 17(1)Read 12 November 2026Insurance distributors “act honestly, fairly and professionally in accordance with the best interests of their customers”.A customer-facing agent answers from approved wording and hands advice to a person.
IDD Art. 20(1)Read 12 November 2026Before a contract, the distributor specifies the customer’s demands and needs from information obtained from the customer.Demands and needs and product advice stay with your licensed staff; the agent can collect the facts.
DORA Art. 2(1)(n), 30Read 12 November 2026Insurance and reinsurance undertakings are in DORA’s scope; Art. 30 sets the ICT contract terms.Our contract carries the Art. 30 terms, and we support your register of information and exit plan.
GDPR Art. 9Read 12 November 2026Health data is a special category; processing it needs one of the conditions in Art. 9(2).A claim that touches health data gets a DPIA and the narrowest data path we can draw.
Our rule. No agent of ours prices or assesses a person for life or health cover outside a full high-risk programme. Claims money always waits for a handler.
Security and privacy, in short

Nine layers around every agent, each with what is partial.

  • Access and identityGoogle sign-in, revocable keys, an agent’s rights narrowed by its spec.Partial No MFA of our own and no enterprise SSO; open sign-up is on.
  • Separation between clientsEvery record and search index filtered by the signed-in identity; a live agent gets its own containers and database.Partial Logical separation in one shared database, not yet tested independently.
  • Hosting, network, encryptionHetzner, Nuremberg; HTTPS with HSTS; no inbound web ports; credentials sealed.Partial Stored data in general is not encrypted at rest.
  • Only the data the task needsThe agent reaches only the systems its approved spec declares; secrets are scrubbed.Partial Personal data is not masked before the model.
  • The model stepProvider set per deployment; each request admitted and recorded before it is sent.Partial The region is named in your contract, not promised EU-only.
  • Untrusted content and testingDocuments are data, never instructions; a safety pack that must all pass.Partial No independent penetration test yet.
  • A person decidesWrites wait for a person; a person approves every release; Stop and spend caps.Partial A live agent’s held actions are approved by our operator, not on your screen.
  • Every run on recordEvery run recorded step by step, with approvals and who decided.Partial Append-only by convention, not tamper-evident; no client export screen.
  • Deletion, export, backupsErasure with a receipt per step; a backup before every release.Partial Export and account deletion by hand, on request; backups not encrypted.
Explore each layer on the trust page
DORA: what we give you as an ICT supplier

The supplier terms a supervisor expects.

If you are in DORA’s scope (Art. 2(1)), you stay fully responsible for your ICT risk, and your supplier contracts carry the Art. 30 terms. We sign them.

  • Register of informationThe facts your register needs for our contract: the service, the functions it supports, our sub-processors and where data is processed (Art. 28(3), Art. 30(2)(a), (b)).
  • Audit and access rightsFor critical or important functions, unrestricted rights of access, inspection and audit for you, an auditor you appoint and your supervisor (Art. 30(3)(e)).
  • An exit planAn exit strategy with a transition period, and the return of your data in an easily accessible format (Art. 28(8), Art. 30(2)(d), 30(3)(f)). Export is by hand today, on request.
  • Incident reporting supportHelp in an ICT incident related to our service at a cost agreed in advance, fast enough for your initial, intermediate and final reports (Art. 19, Art. 30(2)(f)).
Where to start

A timeline up front, a fixed price after one call.

Most teams start with a readiness review of the process they have in mind, or go straight to a pilot when the use case is clearly outside the high-risk list. We don’t publish prices: the proposal after a short scoping call gives a fixed price for your scope, in EUR, excluding VAT.

All engagements and what moves the price
Readiness review · 2–3 weeks

Book a callPrice on request

AI Act and GDPR readiness for one use case.

What’s in it
Pilot agent · 6–10 weeks

Book a callPrice on request

One agent on one process, tested and approved.

What’s in it
Questions

Questions from insurance

Is a claims agent high-risk under the AI Act?

Usually not for property or motor claims. Annex III 5(c) covers risk assessment and pricing of people in life and health insurance, not claims handling in general. GDPR Art. 22 still applies to decisions with significant effects, which is why the agent prepares and a handler decides.

Can an agent recommend a policy to a customer?

We don’t build it to. Under the IDD the distributor specifies the customer’s demands and needs and acts in their best interests, and we keep that with your licensed staff. The agent can collect the facts and answer from approved wording.

Do you handle health data in claims?

Only where the process needs it, with a DPIA, the narrowest data path we can draw and the agent’s access limited to the claim file. Personal data is not masked before the model today, so the data path is designed around that.

How do complaint deadlines work?

The deadline comes from your complaints policy and the rules you are under. The agent shows it on every draft and in the queue; a person sends every reply.

Bring one process. Leave with a risk class, a pass bar and a plan.

A 45-minute scoping call with the people who would build it. You get a first view of the AI Act risk class, the controls it needs, and whether an agent is the right tool at all.

Or read the trust page first