Loss adjuster reports and AI data leakage: what an adjuster’s report gives away in ChatGPT or Claude.

A loss adjuster’s report is written for the insurer, and a broker sees it only when the insurer shares it. It records what the insured said in an interview, what the technician found, what the adjuster thinks the policy does and does not cover, and what the claim is worth. Long and dense, it is exactly what staff paste into ChatGPT or Claude for a summary before a call. This page marks every item in a realistic report that the Privacy Act and the Code of Practice bear on, and shows how the same report is summarised on Ermos inside the brokerage.

The loss adjuster report

Issued to the insurer after the adjuster has inspected the loss and interviewed the insured, and shared with the broker on request.

Tandem Loss Adjusters
Preliminary report to Meridian General · 29 October 2026

Claim.

Instructions.

The insured’s account.

Maintenance.

Technician’s findings.

Power evidence.

Service history.

Comments for the insurer.

Quantum.

An invented report. The serial number is a placeholder. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the brokerage

  1. 1
    The claim and the parties

    The claim number, the insured, the address, the loss date and the broker. A live claim identified to a third party’s servers. Privacy Act guide.

  2. 2
    The insurer’s instructions and the adjuster

    What the insurer asked the adjuster to report on and who did it. The insurer’s own work product, and a named individual. ASIC’s INFO 253 lists loss adjusters among people who often need no licence of their own, and says people who assist an insurer to decide claims may be acting on the insurer’s behalf. The report is the insurer’s document because the insurer instructed it.

  3. 3
    The insured’s account and a private disagreement

    What a director said in interview and a remark about a disagreement with his co-director. Personal information about two directors, and a statement made to the insurer’s adjuster. Privacy Act guide.

  4. 4
    A staff member’s admission

    The practice manager’s statement that the service reminder was put aside. A statement by a named employee that goes to the heart of the decision. Personal information about her. Privacy Act guide.

  5. 5
    The technician’s findings and the serial number

    Technical findings about the failed part and the identifier of the insured machine. Evidence for or against cover.

  6. 6
    The power evidence

    That no outage is recorded and no surge protection is fitted. Evidence on the central factual issue, and a security and safety detail about the premises.

  7. 7
    The service history

    Dates and the manufacturer’s recommended interval. Facts the insurer will rely on and the client’s weakest point.

  8. 8
    Coverage comments

    The adjuster’s view of which policy provisions the insurer can consider, with a note that cover is for the insurer. The insurer’s confidential assessment. Claims handling and settling is a financial service, and the insurer carries the duty to handle claims efficiently, honestly and fairly.

  9. 9
    The repair and income figures and the fees

    The repair quote, the insured’s own estimate of lost profit, flagged as unverified, and the adjuster’s fees. Confidential financial information. Any total needs a person to check it.

  10. 10
    The confidentiality and reliance notice

    The standard notice that the report is for the insurer only. Boilerplate found on any adjuster’s report, with no client data in it.

What happens in Frontier Models (ChatGPT, Claude)

  • The whole document is transmitted to the provider’s servers. On personal plans the user cannot choose where it is stored or processed, and for the major providers that has generally meant outside Australia, under that country’s legal process. Some business plans offer storage in Australia, but the provider still receives the content.
  • On consumer tiers, prompts can be retained and used to train future models.
  • The provider’s staff and systems can reach the content under the provider’s own policies.
  • Once sent, the brokerage no longer controls it. Retention and deletion happen on the provider’s terms.

The same document on Ermos

Ermos Dominion

The document is drafted on a unit in your office, from your own templates and the client file. Every marked item stays on the unit, which keeps working with the internet switched off. Nothing in it is transmitted to an AI provider, so there is no disclosure to analyse: the confidentiality work is the same as for your file server.

Ermos Edge

Your documents, the index and your access boundaries stay on your own instance in an Australian data centre. When a staff member asks for this draft, only their question and the excerpts they are permitted to see go to a named Australian provider, who is contractually barred from keeping the content or training any model on it. Nothing is processed outside Australia.

Frontier Models (ChatGPT, Claude)Ermos EdgeErmos Dominion
Where the document is processedProvider servers, generally offshore on personal plansYour instance in an Australian data centre; answers generated in AustraliaThe unit in your office
Who can read itProvider staff and systems, under policyYour staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothingYour staff only
Is it keptOn the provider’s termsKept on your instance; the provider is barred from retaining itStays on the unit
Can it train a modelOn consumer tiers, yesNo, by contractOnly your own unit learns from it, on your hardware

Part of AI data leakage in insurance brokerages. See how this works in practice: Claims chronologies on Ermos.

Common questions

Do ChatGPT and Claude use the information in a loss adjuster report?
Whatever is pasted or uploaded is sent to the AI company's servers and processed there to produce a reply. As at October 2026, whether a conversation on a personal plan (ChatGPT Free, Plus or Pro; Claude Free, Pro or Max) can be used to train the models depends on a setting each user controls. Business plans (ChatGPT Business and Enterprise, Claude Team and Enterprise) and the developer APIs sit under commercial terms, and OpenAI and Anthropic each say they do not train on that content by default (OpenAI, Anthropic). On either product, the thumbs up and thumbs down buttons can send a whole conversation back for training. A loss adjuster report holds a director's remark about his co-director, the practice manager's admission about the service reminder and the adjuster's comments on cover.
Is it safe to paste a loss adjuster report into ChatGPT or Claude?
No, because the report belongs to the insurer and holds the client's own words. The Office of the Australian Information Commissioner (OAIC) recommends in its guidance on commercially available AI products that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools, because of the significant and complex privacy risks. It records two directors' statements, a named employee's admission, the technician's findings on the scanner's serial number and the adjuster's view of the deterioration exclusion. It is marked as prepared for Meridian General only.
Does deleting the chat remove a loss adjuster report from ChatGPT or Claude?
Not completely, so the user cannot rely on deleting it. Deleting a chat removes it from the user's history, but OpenAI and Anthropic each say they can keep a deleted chat on their systems for up to 30 days, and longer where safety or legal obligations apply (OpenAI, Anthropic). For a report marked confidential to Meridian General, a deleted chat proves nothing about whether the insurer's assessment and the client's interview statements have left the AI company's systems.
How can a brokerage use AI for a loss adjuster report without sending it to ChatGPT or Claude?
Ermos is built so the document stays under your control. On the Dominion AI unit it is drafted inside your office and none of it is sent to an AI provider. On Ermos Edge your documents and index stay on your own instance in an Australian data centre, and only the question and the excerpts the user is permitted to see go to a named Australian provider that is contractually barred from keeping them or training any model on them. A summary is produced from the report on the claim file and the brokerage's own records. The claims handler checks every statement and figure against the report before relying on it in a call.
Does a loss adjuster need a financial services licence?
ASIC's information sheet INFO 253 lists loss assessors and loss adjusters among people who are involved in claims handling and often need no licence of their own, because they are not specifically required to hold one. It describes people who assist an insurer to decide claims through assessment or investigation as people who may also be acting on the insurer's behalf.
Is the client entitled to see the adjuster's report?
That depends on the insurer's obligations, which this page does not cover. A client can ask the insurer for it, and the insurer's written response to a complaint must set out the reasons for rejecting it and the information relied on (ASIC RG 271). Treat any report you hold as the insurer's confidential document unless the insurer says you can share it.
Why does a report the insurer commissioned matter to the brokerage?
The brokerage holds it to help its client, and it contains the insurer's assessment and the client's own interview statements. Sending it to an outside AI tool shares the insurer's work and the client's words with a third party neither agreed to, which sits uneasily with the best interests commitment in the Code of Practice.

See Dominion answering from your own documents.

A demonstration takes 30 minutes, uses no client data, and comes with no obligation.

Book a demo