Artificial intelligence tools are changing the way people research, write, and navigate legal processes. But a recent New Zealand Supreme Court decision is a timely reminder that AI needs careful human oversight, and the consequences of getting it wrong can be serious.

What happened?

In Jones v Family Court at Whangarei [2026] NZSC 1, the Court found that a number of legal authorities cited in submissions by a self-represented litigant appeared to have been "hallucinated" by an AI application, meaning the cases simply did not exist.

The Court was direct about the implications, noting that misuse of AI in legal proceedings has serious consequences for the administration of justice and public confidence in the legal system.

What does this mean in practice?

The Court made clear that everyone filing submissions, whether legally represented or not, is personally responsible for ensuring that all cited authorities are genuine and accurately referenced.

Relying on false citations, including unverified AI-generated material, is not treated as an innocent mistake. In serious cases, it can amount to obstruction of justice or contempt of court.

Why does this matter to you?

AI tools are increasingly accessible, and many people use them to research legal questions, draft documents, or prepare for proceedings. Used well, they can be genuinely helpful. But AI models can, and do, generate plausible-sounding but entirely fabricated case references, and without verification, those errors can find their way into formal submissions.

If you are involved in any legal proceedings, or are considering using AI to assist with legal research or documentation, there are a few important principles to keep in mind:

• Never rely on AI-generated legal citations without independently verifying them. Check that every case exists, says what you think it says, and is still good law.
• AI is a research aid, not a lawyer. It cannot replace professional legal advice, and it carries no accountability if something goes wrong.
• When in doubt, ask. If you're unsure whether something you've prepared is accurate or appropriate to file, seek advice before submitting it.

Our view

At Anthony Harper, we have a considered approach to AI in our practice, including a GenAI policy that sets clear guardrails around acceptable use. Our view is that AI can support certain aspects of legal work, but professional judgment, rigorous verification, and accountability to our clients remain the foundation of everything we do. This decision reflects principles we already hold: that technology should serve the quality of legal advice, not substitute for it.

If you have questions about AI and how it affects your legal matters, please get in touch with one of our experts.