Work Smarter, Not Harder: Top 5 AI Prompts Every Legal Professional in New Zealand Should Use in 2025

New Zealand's legal profession is at a turning point in 2025: the Law Society's nationwide AI research project has mapped how firms and in-house teams are experimenting with generative tools, revealing gaps in confidence, ethics and safe workflows, and the government has now published a national AI strategy to accelerate adoption while emphasising responsible use; that means prompts - the precise instructions lawyers give AI - stop being a novelty and become essential craft.

Well‑written prompts speed work (think summarising a 300‑page disclosure bundle in minutes), reduce hallucinations and help meet Conduct and Client Care obligations, but they also require human oversight, local legal context and privacy safeguards.

For lawyers wanting practical skills, short courses such as Nucamp's AI Essentials for Work bootcamp - practical prompt-writing and workplace AI skills teach prompt design, tool evaluation and governance; pairing that with the Law Society's findings keeps New Zealand practices both efficient and compliant - a national push that turns AI from a risky experiment into a dependable “multi‑tool” for modern practice (Law Society AI research project - AI and Law 2025 webinar, New Zealand national AI strategy analysis - DLA Piper).

“AI is a very powerful multi‑tool in your legal toolbox - learn how to use it, pick it up and put it to work.” - Chris Patterson

Selection started by harvesting high‑value, repeatable legal tasks that matter in New Zealand - contract review, fast due diligence, intake triage and client communications - and then putting candidate prompts through a three‑step vetting loop: design (using the ABCDE framework for precise Audience, Background, Clear instructions, Detailed parameters and Evaluation), iteration (prompt chaining and library templates), and validation (human‑in‑the‑loop review against the “E” criteria plus realistic task tests).

Prompts were adapted for NZ practice by adding jurisdictional cues and Conduct and Client Care safeguards, then stress‑tested on real workflows - e.g., transforming a 50‑page lease into a clause‑by‑clause summary and risk table in minutes, as recommended in ContractPodAi's prompt clinic - while measuring clarity, hallucination risk and speed.

To keep results practical and reproducible, teams kept a prompt library, used prompt‑generation templates from well‑designed prompt libraries, and trialled agentic workflows with mandatory attorney review so AI actions remain assistive, auditable and jurisdiction‑aware (ContractPodAI ABCDE framework and prompt clinic for legal prompts, Thomson Reuters article on prompt libraries for legal AI).

New Zealand case law and practice guidance in 2024–25 make one point crystal clear for prompt‑writers: AI can turbocharge research, but unchecked outputs can slip fake authorities into filings with real consequences.

Recent NZ examples - including Wikeley v Kea Investments Ltd [2024] NZCA 609 and an Employment Court matter where a litigant relied on a non‑existent judgment - show courts will call out AI‑sourced errors and remind users to verify citations; the Law Society has likewise warned practitioners to “beware of fake cases” when using generative tools (New Zealand Law Society guidance on AI tools for legal research – beware of fake cases).

Practical synthesis from commentators stresses existing duties under the Lawyers and Conveyancers Act and the Courts' guidance: checks, transparency and human verification cannot be outsourced (Wilson Harle article on artificial intelligence in the courtroom - implications for New Zealand litigation).

For prompt design that survives scrutiny, build jurisdictional cues, demand sourced citations and iterate until every authority is confirmed - a small procedural habit that prevents the memorable embarrassment of citing cases that don't exist.

“no such case exists, and the plaintiff is reminded that information provided by generative artificial intelligence ought to be checked before being relied on in documents filed in court proceedings”.

Fast, reliable contract review in New Zealand starts with the same fundamentals public procurement guides insist on - verify who the supplier is, check financial resilience, test capacity and capability, and document every step - then use prompts to structure and prioritise those checks so nothing slips through the cracks.

The Government's practical guide on conducting due diligence lays out concrete evidence to ask for (Companies Office checks, audited accounts, credit checks, insurance certificates, references and site visits) and recommends a verification matrix to capture decisions and timing (NZ Government: Conducting due diligence checks).

Lessons from the OAG's review of Callaghan Innovation show why clear scope, documented instructions and transparency matter when third parties are engaged - risks that can be mitigated by making AI outputs auditable and tying them back to primary sources (OAG: Part 2 - The due diligence process).

Pair that with Practical Law's acquisition reminders - know the key risk areas, stay organised, and keep communication open - and prompts can convert a pile of contracts into a ranked risk roadmap that flags, for example, a missing audited account or lapsed insurance certificate before negotiations proceed (Practical Law: 7 reminders for acquisition due diligence).

Drafting demands and client communications for New Zealand work means tailoring tone, substance and remedy to the reader: use a clean, jurisdiction‑specific template as the starting point (for example, Genie AI's New Zealand formal demand letter template can generate a bespoke, editable draft that flags remedies and intended action), make the legal basis and relief requested explicit and anchored in statute or contract law (check the Contract and Commercial Law Act 2017 for cancellation, damages and other contractual remedies), and in specialised areas like trade marks attach the precise evidence and statutory route you're relying on.

Overcoming a citation

IPONZ's guidance explains consent, honest concurrent use and limitation of specifications under section 26.

A tightly focused demand that names the audience (counterparty, agent, or registrar), lists documentary proof and cites the governing statute or IPONZ pathway reads as a clear, auditable instruction rather than a vague threat - think of it as turning foggy risk into a lighthouse beam that makes the next step obvious to everyone involved.

For New Zealand lawyers weighing NZ v Australia differences in 2025, the practical takeaway is jurisdictional clarity in NZ versus doctrinal flux across the Tasman: recent New Zealand developments - from the Court of Appeal's treatment of foreign‑judgment enforcement in Yoonwoo C & C Development v Huh to Wikeley v Kea Investments Ltd and the Contracts of Insurance Act 2024 - sharpen forum, limitation and insurance‑choice issues that make local procedural habits (stay applications, leave to enforce, and forum‑convenience arguments) essential, while Australia continues to wrestle with whether a general duty of good faith should be recognised in commercial contracts (the Journal of Commonwealth Law maps that “crossroads”).

That means prompts and AI workflows used for cross‑border drafting or contract clauses must flag governing‑law, asymmetric jurisdictional terms and consumer protections, and treat each deal like a software build - don't assume an “app” written for one legal system will run unchanged in another (a point Tim Herbert makes about cross‑border contracting).

In practice: add explicit jurisdiction and choice‑of‑law checks to intake prompts, require sourced authority for foreign‑law claims, and surface statutory hooks (like NZ's insurance reforms) so AI suggestions stay tethered to the right forum and rules (Good Faith in Contracts: Australia at the Crossroads, The Conflict of Laws in New Zealand: News and Comment, Mastering Cross‑Border Deals - Tim Herbert Q&A).

“piecemeal, unsettled and unclear”

Intake, timeline and a clear risk roadmap turn messy requests into a predictable, auditable process that New Zealand in‑house teams can actually scale: start with a single “front door” intake form that captures the who/what/when/urgency and integrates with existing tools (Outlook, Teams or SharePoint) so matters auto‑route to the right subject‑matter expert, timelines are set and status dashboards show real progress and KPIs; add AI‑assisted triage to flag high‑risk or time‑critical work, surface statutory hooks or insurance issues, and map a ranked risk roadmap (e.g., missing audited accounts, lapsed insurance, or urgent compliance deadlines) so lawyers can prioritise rather than firefight.

Good intake also creates the data needed for reporting and resourcing - the very metrics Thomson Reuters' intake guide says unlock better decision‑making - while following LawVu's practical design rules (structured forms, meeting the business where it works, and simple self‑service for routine NDAs) to reclaim hours otherwise lost to chasing “quick questions”.

“Implementing technology helped us to achieve several goals, one of which being that we streamlined all of our intakes into one place so that we could keep track of who was asking for what, which then led us to be much more data‑driven so that we could see where are the requests coming from.” - Rosanna Biggs, General Counsel

In short: turning these top prompts into everyday practice means pairing clear ethics, governance and training with the simple discipline of human verification - check every authority, protect client data and document when AI assisted the work - so that a stray “phantom” citation never becomes a courtroom embarrassment.

Rely on the practical supports being rolled out across Aotearoa - the Law Society's AI and Law 2025 programme and workshops give sector‑specific guidance and benchmarking (Law Society AI and Law 2025 webinar), and the Government's new, principles‑based AI Strategy and guidance stress proportionate, risk‑based governance for businesses adopting AI (New Zealand AI Strategy analysis - DLA Piper).

Practical next steps for firms: adopt a written AI use policy, run vendor due diligence and privacy impact assessments, require attorney sign‑off on AI outputs, and train teams on prompt design - or build those skills in a structured programme such as AI Essentials for Work bootcamp - Nucamp, which teaches prompt writing, tool evaluation and workplace governance so prompts help lawyers work smarter without sacrificing professional duties.

"The increasing integration of AI into legal practice is undeniable. Our AI Governance Guide is designed to empower our members. It provides a robust framework for making discerning choices about AI utilisation, always prioritising client confidentiality, data security, and our core professional obligations." - Josh McBride, Richmond Chambers