Form · Client-facing & access-to-justice AI tools
Chatbot transcript sample review worksheet
Working sheet for a lawyer reviewing a sample of client-facing chatbot transcripts for wrong law, advice-line breaches, missed escalations and personal information that should not have been collected or repeated.
- Kind
- Form. A form to complete, or a guide to completing an official one.
- Where it applies
- Australia, New South Wales, Victoria, Queensland, Western Australia
- AI use
- Yellow You and the tool draft together, with you directing. You check every statement you rely on.
- Law stated as at
- 22 October 2026
- Status
- Maintained. Checked on 22 October 2026 against the 4 authorities it cites, for 5 jurisdictions.
- Version
- 1
- Source
- Written for this library
When to use it
Use on a scheduled sample, after a complaint, and before a change to topics, model or channel. It records what the reviewer found. It does not replace the pre-launch checklist or the decision to take the tool offline.
Used for these tasks
- Test a sample of chatbot transcripts for confidential data and wrong law
- Review a client-facing chatbot or intake script against the no-advice line
Each task page says how far AI may help with the task and what the lawyer must review.
How AI may be used
AI may list, from an approved copy of the sample, turns that look like tailored advice, missed crisis or deadline cues, or repeated personal information, for a lawyer to read in the transcript.
AI may
- From an approved copy of the sample, list transcript turns that contain a deadline, a request for advice, a statement of law, or personal information about a third person.
- Count how many conversations in the sample ended with a human, a refusal, or no clear next step.
- Draft the factual description of each flagged turn, quoting the transcript, for the reviewer to check against the original.
AI must not
- Must not mark a statement of law as correct or incorrect.
- Must not decide that a conversation created a lawyer-client relationship, a conflict, or a claim.
- Must not be given transcripts in a tool that is not approved for client information.
The lawyer checks
- The reviewer read the original transcript for every item the tool flagged, and a further random slice the tool did not flag.
- Each statement of law was checked against a current official source before it was recorded as wrong.
- Any conversation that gave tailored advice, missed a safety cue, or leaked information is on the incident report and with the supervising lawyer.
Method: how the Green, Yellow and Red levels are decided. This is guidance for planning, not legal advice.
Read the full text
The full text is available after you log in. It is free. When you log in you can:
- read the whole precedent, with its drafting notes beside it;
- copy it, or download it as a Word file or Markdown;
- add it to your firm’s knowledge base, so colleagues can find it;
- let the AI tools you connect read it from that knowledge base.
The summary, the law it rests on and the AI-use guidance above stay public. Privacy.
More in Client-facing & access-to-justice AI tools
- Client notice: how we use AI on your matter, with consent form
- Client consent to input information into a generative AI tool (United States)
- Terms of use and opening disclaimer for a client-facing legal AI assistant
- Public legal information accuracy review record
- Client consultation when generative AI affects a significant decision (United States)
- Pre-launch assurance and supervision checklist for client-facing AI tools
- No-advice scope and blocked-topics schedule for a client-facing assistant
- No-advice notice for a client-facing tool (United States)
- Collection notice for AI intake and triage tools (APP 5), with short in-chat version
- Client-facing AI incident and near-miss report
- Note escalating a client-facing tool transcript to a lawyer (United States)
- Periodic supervision minute for a client-facing AI tool
- Human escalation and handoff protocol for client-facing AI tools
- Client letter: using AI tools with your legal matter (confidentiality, privilege and evidence)
All precedents for Client-facing & access-to-justice AI tools · Tasks and AI use in this area