Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
A United States insurer has sued the developer of ChatGPT, alleging that the chatbot engaged in the unauthorised practice of law because it was used by a litigant in proceedings against it.
The claim is novel and extraterritorial but still of interest to us. It touches an issue already of direct relevance to many in practice: what happens when an unqualified opposing party treats a general-purpose chatbot as a substitute for a lawyer? Who is responsible for the wasted time and chaos that can ensue?
Nippon Life Insurance Company of America v OpenAI Foundation was filed in the United States District Court, Illinois in March 2026. According to the complaint, a former claimant settled a long-term disability dispute with the insurer in 2024 and signed a release. About a year later, she put her former lawyer’s correspondence recommending acceptance to ChatGPT and asked whether she was being ‘gaslighted’. The chatbot, the complaint alleges, answered in the affirmative, after which she used ChatGPT to prepare litigation seeking to reopen the matter. The Insurer attributes at least 44 filings to the tool.
Nippon Life advances three causes of action against OpenAI: tortious interference with the settlement contract, abuse of process arising from the volume of meritless filings, and the unauthorised practice of law. The third is the point gaining attention.
Notably, the claimant is not a defendant; the insurer has aimed its claim at the developer of the tool rather than the person who used it. OpenAI has said publicly that the complaint lacks merit.
The Queensland approach to regulating unauthoriszed practice has distinct similarities to many of the US states (including Illinois), so authority established there would have some relevance. Of the common-law jurisdictions outside Australia it is most similar in approach to regulating unlicensed practice of law. England and Wales reserve only a short, fixed list of activities to solicitors and leave general legal advice unreserved. New Zealand reserves a narrow set of litigation-related areas and protects professional titles, but likewise does not reserve advice. In those systems, a Nippon-style argument would have little to work with.
Queensland, like most US States, treats legal advice and document drafting as part of the reserved domain, which is the feature the Nippon claim depends on. The Legal Profession Act 2007 (Qld) prohibits “a person” from engaging in legal practice in this jurisdiction unless they are qualified and hold the appropriate practicing certificate.
The LPA offers no closed definition. ‘Engage in legal practice’ is defined only to include ‘practise law’, and whether particular conduct crosses the line is treated as a question of fact, assessed objectively. In Legal Services Commissioner v Walter the Supreme Court restrained an unqualified person who had advised litigants to commence proceedings and had drafted their pleadings and submissions.
Read against that authority the conduct alleged in Nippon Life is not entirely dissimilar. Our prohibition (in contrast to some other Australian States) applies to gratuitous assistance so the fact that a consumer chatbot is free is not, by itself, an answer. A “person” in Queensland legislation generally includes a corporation, so the section is not confined to natural persons.
The unresolved question is whether a prohibition framed around human practitioners can reach an automated system, or the company that builds it. In an increasingly automated world, there are many discussions yet to have about the driver vs car manufacturer vs software liability issue.
Self-represented opponents using AI can impose real cost and delay.
Queensland courts have responded to the issue to some extent by releasing guidance: The Use of Generative AI – Guidelines for Responsible Use by Non-Lawyers. If dealing with a self-represented party furnishing them with a copy of the guideline may be helpful.
Otherwise, the only real lever is specific direction managing the length of affidavits, management of discovery materials and use of AI in response to specific problems presented by an individual.
Supreme Court of Queensland Practice Direction 5 of 2025 – Accuracy of References in Submissions
Queensland Law Society Guidance Statement No 37 – Artificial Intelligence in Legal Practice
Your email address will not be published. Required fields are marked *
The official reopening provided a curtain-raiser to the law school’s 50th anniversary next year.
Isabella Victorsen began with the Brisbane firm in 2023 as a file cadet and was admitted as a solicitor in 2024.
Purcell Taylor Lawyers has announced the promotion of Stephanie Hardman to Senior Associate.
Pro bono lawyer Nareeta Davis will share her experiences about working in a rewarding area of law, at a lunch event in Cairns.
“Getting a fair outcome for a young person, having a magistrate listen and consider, making sure their story is heard. Those are the moments that matter.”
Attendees can enjoy canine company at a stretch session, and hear a judicial officers’ debate, on day two of the Australian Elder Abuse Conference on the Gold Coast.
© Queensland Law Society 2026