Blog

A widely shared survey says legal professionals now think AI verification costs more than it saves. The study’s own authors reach the opposite conclusion — the problem is not the technology but the data and governance underneath it. What that means for New Zealand employers, who have the opposite problem entirely.

New research ranks Kiwi workers lowest of four countries for AI anxiety and guilt. That temperament is a genuine advantage — but a third of the same workforce is passing off AI-generated work without saying so, and the concealment is an employer problem, not a worker one.

Two 2026 redundancy determinations, read together, expose the real test behind consultation: not whether an employer planned ahead — everyone does — but whether the employee still had a genuine chance to change the outcome once consultation opened.

Auckland Council knew about an employee’s personal training business for close to two years before disciplining him for it — and then reached for three internal policy documents that didn’t fit, while the one clause in his employment agreement that did apply never got mentioned. What Solomona v Auckland Council means for any employer sitting on a known issue.

An IRD team leader’s medical incapacity dismissal was found unjustified — not because the decision was wrong, but because the dismissal letter raised new concerns the employee never had a chance to answer. What ZZP v Commissioner of Inland Revenue means for any employer managing a long-term medical absence.

When a vendor withdraws a frontier AI model overnight, the employment law obligations don’t disappear with it. What the Fable 5 and Mythos 5 shutdown means for New Zealand employers who have integrated AI into live business processes.

Three days after the head of a leading AI laboratory argued that governments should be able to block frontier models, a government blocked his. Here is what that week revealed about who actually holds the door — and what it means for New Zealand employers and firms.

The Lex Praxis AI Output Auditor is now live — a free tool that reads AI-generated employment law text and classifies every claim as Verified, Needs Checking, Unsupported, or Likely Wrong. It will tell you more about what your AI produced than any amount of re-reading or prompt engineering the text yourself.

New Zealand's Supreme Court has confirmed that AI-generated output is your responsibility the moment you use it. In Jones v Family Court at Whangārei [2026] NZSC 1, the Court warned that reliance on unverified AI output may in serious cases amount to obstruction of justice. That principle reaches well beyond the courtroom.

The NZ Law Commission has been asked to review automated decision-making by government. New Zealand currently has no overarching law, standards, or guidance addressing how agencies should use AI in a legally compliant and consistent manner. The gap that acknowledgement reveals does not belong only to the public sector.

The Employment Relations Amendment Act 2026 is the most significant change to NZ employment law since 2018. It rewrites contractor status, removes dismissal protections for high earners, and restructures personal grievance remedies. It does not mention artificial intelligence once. On what that silence means — and what it will cost.

Every AI hallucination in a legal filing has a human signature on it. A lawyer filed it. The failures we keep attributing to AI are, on closer examination, failures of professional judgement. On what it means to stay in the driver's seat — and what happens when you don't.

A UK tribunal has ruled that uploading privileged client material to a public AI platform destroys privilege permanently. NZ courts have already seen six AI filing failures since 2024. Here is what the case establishes, why it will travel, and what NZ firms need to do before our courts get there on privilege.

A health issue. A gap. A return to practice with different eyes. And the realisation that AI had moved from a theoretical future to a present-tense reality — while most of the legal world wasn't paying attention. This is why Lex Praxis exists.

When you decide to build a business at the intersection of employment law and artificial intelligence, the first practical question isn't about positioning or pricing or even clients. It's about whether you can actually use the tools you're planning to sell. Here's what happened when I tried.