Law Firm AML/CTF Customer Identification Records: Use case and AI Confidentiality
From 1 July 2026, law firms that provide certain services, such as helping a client buy real estate, must establish who the client is and assess the money laundering risk before they start, with a limited allowance in conveyancing to finish the check after exchange or before settlement. The record of that check is a form: names, dates of birth, identity document numbers, where the money came from and a risk rating. Staff paste it into ChatGPT or Claude to tidy a note or fill a template. This page marks every item in a customer identification record that rule 9, the Privacy Act and the AML/CTF rules bear on, and shows how the same record is completed on Ermos inside the firm.
The AML/CTF customer identification record
Completed when a client engages the firm for a service the AML/CTF Act regulates, and kept for seven years after the relationship ends.
Customer identification record
Carrick Doyle Lawyers · AML/CTF program · initial customer due diligence
| Matter | OKAF-0399 · item 1 designated service: assisting with the purchase of 41 Pittwater Road, Manly NSW 2095; settlement 30 October 2026; balance on settlement $1,642,118.4010 |
| Customers | Daniel James Okafor, born 14 June 1984; Priya Anjali Okafor, born 2 February 1987; 9 Wattle Street, Balgowlah NSW 20931 |
| Identity documents | Daniel: Australian passport PA1234567. Priya: NSW driver licence 1234 5678.2 |
| Verification | Both documents checked through the document verification service on 11 September 2026: match. Document details recorded here. Verified by JM.3 |
| Screening | Politically exposed person and sanctions screening, 11 September 2026: no match for either customer.4 |
| Nature and purpose | Purchase of a home to live in; one transaction; customers instruct jointly. |
| Source of funds | Contribution of $362,118.40: $338,000.00 net proceeds of the sale of 6/20 Dee Why Parade (settled 4 September) and $24,118.40 from Priya’s savings. Occupations: Daniel, civil engineer; Priya, occupational therapist.5 |
| Finance | Westpac loan of $1,280,000, approved 2 September 2026.6 |
| Risk rating | Low. Two resident individuals, no foreign links, funds from a property sale and a regulated lender. Enhanced due diligence not triggered. Next review at settlement.7 |
| Timing | Completed 11 September 2026, before contracts were exchanged on 18 September. Delayed initial CDD not used. |
| Retention | Keep for seven years from the end of the relationship, and not before 30 October 2033.8 |
| Completed and approved | JM, paralegal, 11 September 2026. Approved by Eleanor Vance, partner.9 |
An invented form. The passport and licence numbers are placeholders. Hover or tap any highlighted item for the analysis; the full list is below.
Client identification record pre-filler
An Agent+ that completes the first pass of the customer identification form from information the firm already holds: names, document details, the service sought and the source of funds as the client described it. It works from your own form and procedure, and marks every blank or mismatch.
What it does
- Reads the engagement details and client documents on file
- Pre-fills your identification form from that information
- Flags blanks, mismatches and items needing a second look
What stays with your people. Your AML/CTF compliance officer or responsible partner verifies each entry, assigns the risk rating and signs the record.
An Agent+ is built once and anyone in the practice can run it. It works from your own documents and runs on Ermos Dominion, the AI unit in your office, or on Ermos Edge, your own instance in an Australian data centre.
What leaves the firm
- 1The customers, their birth dates and their address
Two named people with dates of birth and a home address. Enough to open accounts in their names. Personal information under the Privacy Act, and confidential to the clients under rule 9. Rule 9 confidentiality guide.
Frontier Models (ChatGPT, Claude) Transmitted to the provider, generally offshore on personal plans Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 2Passport and licence numbers
Two government identifiers. The Law Society of NSW’s AML/CTF guide says a practice need not copy identification documents for AML/CTF purposes but must record the details it used, such as the passport details, so this field is where the numbers live. The material identity theft is built from. Privacy Act guide.
Frontier Models (ChatGPT, Claude) Two government identifiers transmitted Ermos Edge Leaves only if the numbers are inside the question; to a named Australian provider that keeps nothing Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 3The verification result
The steps taken to verify the customers and the outcome. Records of what customer information was collected, the steps taken to verify it and the decision about the level of due diligence are the records the guide says a practice must make and keep. Privacy Act guide.
Frontier Models (ChatGPT, Claude) The verification steps and result transmitted Ermos Edge The verification service runs the check; the AI does not Ermos Dominion The verification service runs the check; the AI does not Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 4The screening result
Whether either customer is a politically exposed person or designated for targeted financial sanctions. Initial customer due diligence requires the practice to establish those facts on reasonable grounds, so the result is part of the record. Privacy Act guide.
Frontier Models (ChatGPT, Claude) Screening results about named people transmitted Ermos Edge The screening tool runs the check; the AI does not Ermos Dominion The screening tool runs the check; the AI does not Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 5Where the money came from, with the amounts
A sale of a named property, the net proceeds, a partner’s savings and both occupations. Recording the source of funds and wealth is required where enhanced due diligence applies, and this firm records it for property purchases anyway. Personal financial information, and the same figure as the settlement email. Rule 9 confidentiality guide.
Frontier Models (ChatGPT, Claude) The source of funds and the amounts transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 6The loan
A lender, an amount and an approval date. Financial information about the clients and a fact a payment redirection fraud would use. Rule 9 confidentiality guide.
Frontier Models (ChatGPT, Claude) A lender, an amount and an approval date transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 7The risk rating and its basis
The firm’s own judgement of how closely it looked at these clients and why. The firm’s AML/CTF control material, which a person trying to avoid detection would like to read. Law Society AI guidance.
Frontier Models (ChatGPT, Claude) The firm’s internal risk judgement transmitted Ermos Edge The AI can draft a risk rating but cannot own the judgement; a person must decide Ermos Dominion Stays on the unit, but the AI cannot own the judgement; a person must decide Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 8The retention rule
A statement that records are kept for seven years from the end of the relationship, which the guide gives as the period for customer due diligence records. A public rule with a date attached.
Frontier Models (ChatGPT, Claude) A public rule with no client detail beyond a date; low value alone Ermos Edge A person must check the retention period against the AML/CTF rules Ermos Dominion A person must check the retention period against the AML/CTF rules Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 9The staff
Initials and a partner’s approval that tie the record to the people responsible for the firm’s AML/CTF program. Supervising AI guide.
Frontier Models (ChatGPT, Claude) Transmitted, and the leak is traceable to you Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case - 10The matter, the property and the balance
A specific address, a settlement date and the amount moving. The facts a payment redirection fraud needs, in the first line of a compliance form. Rule 9 confidentiality guide.
Frontier Models (ChatGPT, Claude) A live transaction, its date and the amount transmitted; on consumer tiers it can train future models Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Client identification record pre-filler. Pre-fills the customer identification record from the engagement details and documents already on file, and flags missing or inconsistent entries for the responsible person. See the use case
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 firm 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 matter 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 Edge | Ermos Dominion | |
|---|---|---|---|
| Where the document is processed | Provider servers, generally offshore on personal plans | Your instance in an Australian data centre; answers generated in Australia | The unit in your office |
| Who can read it | Provider staff and systems, under policy | Your staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothing | Your staff only |
| Is it kept | On the provider’s terms | Kept on your instance; the provider is barred from retaining it | Stays on the unit |
| Can it train a model | On consumer tiers, yes | No, by contract | Only your own unit learns from it, on your hardware |
Part of Use cases in law firms.
Common questions
Do ChatGPT and Claude use the information in an AML/CTF customer identification record?
Is it safe to paste an AML/CTF customer identification record into ChatGPT or Claude if we take the names out?
Does deleting the chat remove an AML/CTF customer identification record from ChatGPT or Claude?
How can a law firm use AI for an AML/CTF customer identification record without sending it to ChatGPT or Claude?
Which legal work is caught by the AML/CTF rules from 1 July 2026?
Can a law firm start a conveyancing matter before identifying the buyer?
Does a law firm have to keep copies of identity documents for AML/CTF purposes?
See Dominion answering from your own documents.
A demonstration takes 30 minutes, uses no client data, and comes with no obligation.
Book a demo