AI can take a problem from zero to 100 very quickly
Across New Zealand's rental landscape, a quiet technological shift is straining the institutions built to keep landlords and tenants in balance. Armed with AI tools like ChatGPT, tenants are transforming minor maintenance grievances into elaborate legal documents — not necessarily because their cases are stronger, but because the tools they use are designed to amplify whatever they are given. The Tenancy Tribunal, already stretched, is absorbing the consequences in the form of longer delays and more complex filings, raising a question as old as access to justice itself: does giving people more powerful tools make the system more fair, or simply more congested?
- What once resolved with a phone call now arrives as a five-page legal complaint, and property managers say the pace of escalation is unlike anything they have seen before.
- Eight in ten Bayleys property managers report tenants using AI to draft disputes, with half saying these interactions have become significantly harder to navigate.
- Hearing wait times at the Tenancy Tribunal have ballooned from three to four weeks to two full months, leaving landlords accumulating unpaid rent debt while cases sit in an overwhelmed queue.
- Industry leaders warn the ripple effect is financial: more complex disputes mean more staff, higher management fees, and ultimately higher rents for the very tenants the tools were meant to empower.
- The Tenancy Tribunal is drafting guidance on AI use in filings — not to ban the tools, but to clarify that an amplifier is not the same thing as a lawyer.
Property managers across New Zealand are watching a familiar problem become something harder to solve. A leaking tap used to mean a quick call and a repair visit. Now, tenants armed with ChatGPT are turning that same tap into a five-page legal complaint, complete with citations to tenancy law that may or may not apply. Eight in ten property managers at Bayleys report tenants using AI to draft complaints, six in ten say routine issues are becoming full disputes, and half say the conversations themselves have grown harder to manage.
Will Alexander of Bayleys describes the dynamic with precision: AI responds to the questions you feed it, and those questions tend to be one-sided. A tenant frustrated about a minor issue will phrase their query in a way that invites validation — and the AI obliges. What emerges rarely resembles the original problem. Klim Andreev of Keen on Rentals puts it plainly: AI gives tenants false hope by mirroring their frustration and telling them what they want to hear.
The downstream consequences are real. Duncan McLean of Property Scouts warns that if firms must hire additional staff to manage the volume of AI-generated disputes, management fees will rise — and rents will follow. Meanwhile, the Tenancy Tribunal is straining under the load. Hearings that once took three to four weeks to schedule now take two months, leaving landlords in debt limbo while cases queue.
Not everyone reads this as a crisis. Sarina Gibbon of Tenancy Advisory argues AI may help renters navigate a system that can feel opaque and intimidating. But she is careful to call it an amplifier, not an equalizer — one that strengthens strong cases and exposes weak ones with equal indifference. The Tribunal is now preparing guidance on AI use in filings, not to prohibit the tools, but to remind both sides that amplifying a grievance is not the same as resolving one — and that sometimes the fastest path forward is still the oldest: picking up the phone.
Property managers across New Zealand are watching a familiar problem transform into something harder to solve. A leaking tap used to warrant a quick phone call and a repair visit. Now, tenants armed with ChatGPT are turning that same tap into a five-page legal complaint, complete with citations to tenancy law that may or may not apply to their situation.
The shift is real enough that it's starting to reshape how the rental system works. Eight in ten property managers at Bayleys report that tenants are now using AI tools to draft complaints and explore their rights. Six in ten say this is converting routine issues into full-blown disputes. Half say the conversations themselves have become harder to manage. The numbers suggest this isn't a fringe phenomenon—it's becoming standard practice.
Will Alexander, head of innovation and property management at Bayleys, describes what's happening with a phrase that captures the speed of escalation: AI can take a problem from "zero to 100" very quickly. The issue, he explains, is that AI responds to the questions you feed it, and those questions tend to be subjective and one-sided. A tenant frustrated about a minor maintenance issue will phrase their question in a way that invites the AI to validate their frustration, and the AI obliges. What emerges is a complaint that bears little resemblance to the original problem.
Duncan McLean, national director of Property Scouts, sees a downstream consequence that could touch every renter in the country. If property management firms need to hire additional staff to handle the volume and complexity of AI-generated disputes, management fees will rise. And when management fees rise, rents follow. The concern isn't theoretical—some tenants are already using AI to explore legal threats, even when the legal advice the AI provides is inaccurate. Klim Andreev, a property manager at Keen on Rentals, describes the effect bluntly: AI gives tenants false hope by mimicking their frustration and telling them what they want to hear.
The Tenancy Tribunal, the institution designed to resolve these disputes, is feeling the strain. Jason Waugh, general manager of Lodge City Rentals, says he's never seen delays like the ones happening now. A simple application for unpaid rent used to take three to four weeks to reach a hearing. Now it takes two months. For a landlord, that means two months of debt accumulating while the case sits in a queue. The Tribunal itself has noticed the pattern—filings that use AI-generated material are longer, more complex, and harder to parse.
Not everyone sees this as purely negative. Sarina Gibbon, director of Tenancy Advisory, argues that AI may be helping renters understand their rights in a system that can feel intimidating and opaque. The problem, she says, is that AI is an amplifier, not an equalizer. If you have a strong case and understand how tenancy law applies to your situation, AI can strengthen your position. But if your case is weak and your understanding of the law is thin, AI will amplify that weakness just as readily. David Faulkner from Property Brokers doesn't object to tenants using AI—he objects to the system-wide slowdown it's creating. "This is not about tenants exercising their rights," he says. "This is about slowing down the whole system, which helps no-one."
The Tenancy Tribunal is now preparing guidance on how AI should be used in tenancy disputes. The guidance won't ban the tools—it can't, and probably shouldn't. But it may help both tenants and property managers understand that a tool designed to amplify whatever you feed it isn't the same as legal advice, and that sometimes the fastest way to solve a problem is still the oldest one: picking up the phone.
Notable Quotes
AI can take a problem from zero to 100 very quickly— Will Alexander, Bayleys head of innovation and property management
AI is an amplifier, not an equalizer—it strengthens strong cases and weak ones alike— Sarina Gibbon, Tenancy Advisory director
The Hearth Conversation Another angle on the story
Why does it matter if a complaint is five pages instead of one sentence? Isn't more detail better?
Only if the detail is accurate and relevant. When AI generates a complaint, it's often padding the original grievance with legal language and worst-case scenarios the tenant never actually intended to raise. That creates work for property managers to untangle what the real issue is, and it clogs the Tribunal with cases that take longer to resolve.
But doesn't AI help tenants who don't know their rights?
It can, but that's not what's happening most of the time. Gibbon's point about amplification is crucial—AI doesn't teach you the law, it just makes whatever you already believe sound more authoritative. A tenant with a legitimate complaint gets a stronger voice. A tenant with a weak case gets a more convincing-sounding weak case.
So the real problem is that the system is getting slower?
That's part of it. But there's something deeper: AI is changing the tone of the conversation. What used to be a negotiation between two parties is becoming a legal standoff, even when the original problem was minor. That makes everything more adversarial.
Will property managers just refuse to use AI-generated complaints?
They can't refuse them. But they're already talking about raising fees to cover the extra work. That cost gets passed to tenants through higher rents. So ironically, tenants using AI to protect themselves might end up paying more for housing.
Is the Tribunal's guidance going to fix this?
Probably not entirely. Guidance can clarify what AI should and shouldn't be used for, but it won't stop people from using it. The real fix would be cultural—tenants and landlords both understanding that not every problem needs to become a legal case.