Sometimes, and it depends on who is asking. Australian lawyers must disclose AI use to a court when that court's practice note requires it. The Federal Court's GPN-AI requires disclosure when generative AI shapes evidence, and NSW courts require affidavits to state that generative AI was not used to generate them. No rule requires lawyers to tell every client about every AI tool, but regulators expect firms to be transparent about AI use, bill accurately and protect confidentiality. From 10 December 2026, the Privacy Act also requires some firms to explain automated decision-making in their privacy policy.
AI transparency is now a practical question for every Australian law firm. Most firms already use AI somewhere, whether it is a research assistant, a drafting tool or a feature built into software they already pay for. The question has moved from "can we use AI?" to "who do we have to tell, and what do we have to say?"
The answer depends on the audience: the court, the client or the regulator. This guide sets out the AI transparency rules that apply to Australian law firms as of October 2026, with a policy checklist and a sample engagement letter clause.
This guide is general information, not legal advice. Always check the current practice notes of each court you appear in.
What does AI transparency mean for law firms?
AI transparency means being open about when, how and why AI tools are used in legal work, so the people who rely on that work can trust it. For Australian law firms, it runs in three directions:
- Courts: disclosing AI use in court documents where a practice note or direction requires it, and verifying everything you file.
- Clients: explaining how AI is used on their matters, how their information is protected and how it affects fees.
- The public and regulators: describing automated decision-making in your privacy policy, where the Privacy Act applies.
The common thread is accountability. Every Australian court and regulator that has spoken on AI has taken the same position: AI does not change a lawyer's duties. The lawyer, not the tool, is responsible for the work.
Do Australian lawyers have to disclose AI use to the courts?
Yes, in some circumstances, and the rules differ by court. There is no single national rule, so if you practise across jurisdictions you need a forum-by-forum approach. For a deeper look at how professional duties interact with these rules, read our guide to the Legal Profession Uniform Law, AI and court practice notes.
Federal Court of Australia: GPN-AI (April 2026)
The Federal Court published its Use of Generative Artificial Intelligence Practice Note (GPN-AI) on 16 April 2026. It applies to everyone who appears before or files documents with the Court, including self-represented litigants, witnesses and people producing documents under subpoena.
Disclosure is mandatory where generative AI was used to:
- summarise or analyse information a witness relies on to state a fact or give an opinion
- create images, video, sound recordings or other multimedia presented to the Court
- do anything else that might reasonably be expected to affect the admissibility of evidence, or how the Court uses it.
The disclosure goes in the body of the document, at the start, and should briefly say where and how generative AI was used. The Court can also order disclosure in other circumstances, and expects everyone to be able to explain which AI tool they used, how and for what purpose.
Pleadings and submissions do not need a disclosure statement under GPN-AI, but the person responsible for the document must have confirmed that cited authorities exist and support the proposition, that evidence referred to exists, and that chronologies are accurate. Consequences for getting it wrong include adverse costs orders.
NSW Supreme and District Courts: Practice Note SC Gen 23
NSW takes the strictest approach in Australia. Practice Note SC Gen 23 took effect on 3 February 2025, and the District Court has adopted it.
- Generative AI must not be used to generate the content of affidavits, witness statements or character references, and each must state that generative AI was not used to generate it.
- Generative AI must not be used to draft or prepare an expert report without the Court's prior leave.
- Where generative AI helps prepare written submissions, the author must verify that every citation exists, is accurate and is relevant.
Victoria: Supreme and County Court guidelines
The Supreme Court of Victoria was the first Australian court to publish AI guidelines for litigants, in May 2024, and the County Court has similar guidelines. They expect parties and practitioners to disclose AI assistance to each other and, where appropriate, to the Court. They are guidelines rather than a mandatory practice note, but practitioners should treat them as the Court's expectations.
Queensland: Practice Direction 5 of 2025
The Supreme Court of Queensland's Practice Direction 5 of 2025, issued on 24 September 2025, does not require a disclosure statement. Instead, the person responsible for submissions must be identified and must verify every reference to legislation, case law and other sources, in written and oral submissions. Other Queensland courts and tribunals have adopted the same approach.
Court AI disclosure rules at a glance
Federal Court of Australia: Under GPN-AI (16 April 2026), disclosure is required where AI use affects evidence, and whenever the Court directs it. Disclose at the start of the document, and verify all authorities and evidence.
NSW Supreme and District Courts: Under Practice Note SC Gen 23 (3 February 2025), affidavits, witness statements and character references must state that generative AI was not used to generate them. Generative AI cannot be used in evidence, and expert reports need the Court’s leave.
Victorian Supreme and County Courts: Under the Guidelines for litigants (2024), disclosure is expected but not mandated. Disclose AI use to other parties and the Court where appropriate.
Queensland courts: Under Practice Direction 5 of 2025, there is no general disclosure statement. Every reference in submissions must be verified.
Do lawyers have to tell clients they are using AI?
No rule in the Australian Solicitors' Conduct Rules (ASCR) requires a firm to disclose every AI tool on every matter. But your existing duties make transparency with clients the safe default.
In December 2024, the Law Society of NSW, the Legal Practice Board of Western Australia and the Victorian Legal Services Board + Commissioner issued a joint Statement on the use of artificial intelligence in Australian legal practice. It applies to NSW solicitors and to solicitors and barristers in Victoria and Western Australia. For client relationships, it comes down to four expectations:
- Be transparent. Tell clients how AI is used on their work, and address any ethical concerns they raise proactively.
- Protect confidentiality. Your duty of confidentiality (ASCR rule 9) applies to anything you put into an AI tool. Public tools may store or process client data offshore.
- Bill honestly. Time and work items billed must accurately reflect the legal work done by people in your firm. You cannot bill hours for work an AI tool did in minutes.
- Stay competent. Your duty to deliver legal services competently and diligently (ASCR rule 4.1.3) means understanding each tool's limits and checking its output.
Costs disclosure under the Legal Profession Uniform Law matters too. If AI changes how you charge, for example through fixed fees or a technology charge, that belongs in your costs disclosure.
In practice, most SME firms handle client transparency with three things: a short clause in the engagement letter, an AI statement on the firm's website, and a conversation on the matter when the client asks or the stakes are high.
The Privacy Act change coming on 10 December 2026
From 10 December 2026, new Australian Privacy Principles 1.7 to 1.9 require APP entities to explain automated decision-making in their privacy policies. The obligation applies when a computer program uses personal information to make a decision, or to do something substantially and directly related to making one, and the decision could reasonably be expected to significantly affect someone's rights or interests.
Where it applies, your privacy policy must describe:
- the kinds of personal information used in automated decision-making
- the kinds of decisions made solely by a computer program
- the kinds of decisions where a program plays a substantial and direct role.
For law firms, the most likely trigger is automated client intake, triage or screening, such as a system that declines enquiries or routes matters without a person reviewing them. Drafting and research tools whose output a lawyer reviews are less likely to be in scope, but map your workflows before you decide. The Privacy Act generally exempts businesses with annual turnover of $3 million or less, so first confirm whether your firm is an APP entity.
How to build an AI transparency policy for your law firm
A written AI policy is the simplest way to meet court, client and regulator expectations at once. Seven steps:
- List your AI tools and what each is approved for, including AI features built into software you already use.
- Classify tasks by risk. The joint statement recommends limiting AI to lower-risk tasks that are easier to verify. Evidence and expert material are the highest risk.
- Set data rules. Decide which information may go into which tool. Confidential, privileged and suppressed information should only go into tools with contractual, onshore data protection, and even then, consider the implied undertaking on documents obtained in litigation.
- Require verification. A lawyer checks every citation, quote and factual statement produced with AI before it leaves the firm.
- Map court rules for each forum you appear in, and build the required disclosure wording into your precedents.
- Update client documents: engagement letters, costs disclosures, your website AI statement and, where needed, your privacy policy.
- Keep records and review. Record which tool was used for what on each matter, so you can answer if a court asks. Review the policy at least every six months.
Sample AI clause for client engagement letters
We may use secure artificial intelligence tools to help with research, drafting and document review on your matter. These tools are hosted in Australia and do not use your information to train AI models. A lawyer reviews all work produced with their help, and we remain fully responsible for the advice we give you. We only charge for work our people actually do. Please tell us if you would prefer that we do not use AI on your matter.
Adapt the clause to your tools and costs agreement, and only state that your tools are hosted in Australia if they are.
How private legal AI makes AI transparency easier
Transparency is easier when you can answer three questions with confidence: where client data goes, whether it is kept, and how outputs can be checked. Public chatbots make all three hard to answer.
LegalScout was built around those questions:
- Onshore hosting. All data stays in AWS Sydney (ap-southeast-2). See Australian cloud hosting.
- Zero data retention. Model providers retain none of your data and never train on it.
- Cited answers. Ask Scout answers Australian legal research questions with citations you can verify before anything reaches a client or a court.
- Aligned to the Australian Privacy Principles. See our approach to Privacy Act compliance, or share the Trust Centre with clients who ask how their data is protected.
The Federal Court notes that closed AI tools carry a lower risk of inadvertent disclosure than public ones, while still expecting firms to protect confidential information rigorously. A private platform gives you a stronger starting point. Your policy and verification steps do the rest.
Try it on your own documents. Start free with LegalScout, no credit card required, or explore AI legal research for Australian lawyers.
Sources
- Federal Court of Australia, Use of Generative Artificial Intelligence Practice Note (GPN-AI), 16 April 2026
- Supreme Court of NSW, Practice Note SC Gen 23 – Use of Generative Artificial Intelligence, commenced 3 February 2025
- Supreme Court of Victoria, Guidelines for litigants: responsible use of artificial intelligence in litigation, May 2024
- Supreme Court of Queensland, Practice Direction 5 of 2025 – Accuracy of references in submissions, 24 September 2025
- Victorian Legal Services Board + Commissioner, Statement on the use of artificial intelligence in Australian legal practice, December 2024
- Gilbert + Tobin, Automated decision-making transparency under the Privacy Act
About LegalScout
LegalScout is a private legal AI platform built by Australian lawyers for SME law firms. Hosted entirely in AWS Sydney and aligned to the Privacy Act 1988 (Cth), LegalScout combines intelligent legal research, document drafting, contract review, and financial modelling into a single credit-based subscription - with no per-seat licensing. Book a demo to see how it works with your own documents.
Frequently asked questions
Do Australian lawyers have to tell clients when they use AI?
No rule requires disclosure of every AI tool on every matter. However, the joint statement by the Law Society of NSW, the Legal Practice Board of WA and the Victorian Legal Services Board + Commissioner expects lawyers to be transparent with clients about AI use, protect confidentiality and bill only for work actually done. Most firms add an AI clause to their engagement letters.
Do lawyers have to disclose AI use to Australian courts?
It depends on the court. The Federal Court's GPN-AI requires disclosure at the start of a document when generative AI is used in ways that affect evidence. In NSW, affidavits, witness statements and character references must state that generative AI was not used to generate them. Victoria expects disclosure, and Queensland requires every reference in submissions to be verified.
Can lawyers use ChatGPT for client work in Australia?
Lawyers can use general AI tools, but courts and regulators warn against entering confidential or privileged information into public tools, which may store or process data offshore. Any output must be verified before it is relied on. Most firms use a private legal AI platform with onshore hosting and zero data retention for client matters.
What changes for law firms under the Privacy Act on 10 December 2026?
Firms covered by the Privacy Act must describe automated decision-making in their privacy policies when software makes, or substantially and directly contributes to, decisions that could significantly affect individuals. Automated client intake and triage are the most likely triggers for law firms.