Firm-wide obligation
Client-facing & access-to-justice AI toolsClient-facing and access-to-justice AI tools
Define the service boundary, supervision and escalation route before publishing a client-facing tool. Check its outputs, data handling and accessibility against the actual users and intended purpose.
Key facts
- Tasks covered
- 5
- Tasks with a step-by-step checklist
- 1
- Local differences
- 2 jurisdictions where this area changes
Jurisdiction
Tasks
GreenLawyer reviewYellowClose supervisionRedLawyer-led
| Task | How AI is used | Human involvement | What the lawyer must review |
|---|---|---|---|
| Review a client-facing chatbot or intake script against the no-advice line | From the script and sample transcripts, list answers that look like tailored advice, missing escalations and data-handling gaps. | RedLawyer-led |
|
| Deploy a US client-facing chatbot without it practising law unsupervised | Draft proposed user-facing disclosures (not a lawyer, not legal advice, how data is used) from the firm’s policy. | RedLawyer-led |
|
| Deploy a client-facing intake tool under the Conduct and Client Care Rules | Draft a description of what the tool will ask and what it will not decide. | RedLawyer-led |
|
| Review public self-help content for accuracy and the advice line | Compare the draft page to the current sources the lawyer has named, list stale statements and flag wording that reads as tailored advice. | RedLawyer-led |
|
| Build an escalation matrix for a client-facing tool | From the tool flows, list question types, the point a human must take over, and the record that must be kept. | YellowClose supervision |
|
| Draft client-facing disclosures that the firm uses AI on the matter | Produce first-draft retainer and website language from the professional-guidance headings the lawyer has listed, limited to how this firm actually uses tools. | YellowClose supervision |
|
| Test a sample of chatbot transcripts for confidential data and wrong law | From a permitted sample, list transcripts that include identifying facts, hallucinated rules or missed escalations. | YellowClose supervision |
|
What changes in New Zealand
| Topic | What changes |
|---|---|
| Responsibility | The Law Society lists client-facing chatbots as a use and a risk. You remain responsible for the legal service. |
| Client care | Say who will do the work and how (r 3.5). Consult the client on steps to carry out instructions (r 7.1). |
| Privacy | Plain-language transparency, particular care with children and human review before acting. Do not test a chatbot on client or personal information. |
| Limits on AI assistance |
|
- NZLS: Generative AI guidance for lawyers (March 2024) · 2026-09-23
- Lawyers and Conveyancers Act (Lawyers: Conduct and Client Care) Rules 2008 · 2026-09-23
- OPC: Generative artificial intelligence (15 June 2023) · 2026-09-23
- OPC: Artificial Intelligence and the Information Privacy Principles · 2026-09-23
What changes in United States
Known here as: Client-facing and access-to-justice AI tools
| Topic | What changes |
|---|---|
| Ethics | Formal Opinion 512 and Florida 24-1 treat public chatbots as advertising, confidentiality and unauthorised-practice risks. NYC Bar 2024-5 flags advertising and discrimination. |
| Agentic systems | California's 2026 guidance: no agentic legal advice or representative acts without meaningful lawyer supervision. |
| Colorado | HB26-1263 (signed 1 July 2026, in force 1 January 2027): chatbot outputs must not be represented as equivalent to licensed professional services. |
| Limits on AI assistance |
|
- ABA Formal Opinion 512, Generative Artificial Intelligence Tools (29 July 2024) · 2026-09-23
- Florida Bar Ethics Opinion 24-1 · 2026-09-23
- State Bar of California, Practical Guidance for the Use of Generative Artificial Intelligence in the Practice of Law (2026) · 2026-09-23
- Colorado Attorney General: Automated Decision-Making Technology and Chatbot Safety rulemaking · 2026-09-23
- New York City Bar Formal Opinion 2024-5: Generative AI in the Practice of Law · 2026-09-23
Precedents for this area
- Client notice: how we use AI on your matter, with consent form (Notice, AI use Yellow). Plain-English notice telling clients which tasks the firm uses AI for, what it never uses AI for, how their information is protected and charged for, and how to consent, restrict or refuse.
- Terms of use and opening disclaimer for a client-facing legal AI assistant (Agreement, AI use Yellow). Website terms of use and the opening chat message for a law practice's AI assistant: it is AI, gives general information only, creates no lawyer-client relationship, handles urgent matters and data safely, and preserves consumer rights.
- Pre-launch assurance and supervision checklist for client-facing AI tools (Checklist, AI use Red). Go-live checklist for chatbots, intake assistants and self-help tools: scope and advice line, testing on real enquiries, escalation, privacy and security, consumer law, accessibility, named supervising lawyer and monitoring.
- Collection notice for AI intake and triage tools (APP 5), with short in-chat version (Notice, AI use Yellow). APP 5 collection notice for an AI chatbot or online intake form, covering what is collected, why, who processes it (including offshore AI providers), how enquiries are triaged, and whether any decision is automated.
- Human escalation and handoff protocol for client-facing AI tools (Policy, AI use Yellow). Protocol listing the triggers on which a client-facing AI tool must stop and hand over to a person (urgency, deadlines, family violence, self-harm risk, vulnerability, conflicts, requests for advice), with response times and records.
- Client letter: using AI tools with your legal matter (confidentiality, privilege and evidence) (Letter, AI use Yellow). Letter warning clients that putting legal advice or matter details into AI tools can risk confidentiality and privilege, that evidence must be in their own words, and that AI legal information can be wrong.
All precedents for Client-facing & access-to-justice AI tools. Signed-in accounts can read the full text.