Tenants, ChatGPT and the Private Rented Sector: A trend landlords can’t ignore
Nikki Lewis, Director of Operations at CPS Homes, brings us his observations on the rise of AI in tenant communications and what it means for landlords…
I’ve been thinking about writing an observational piece on the increasing use of so-called “AI assistants” in the private rented sector for some time now. A particular email I received from a tenant recently was, to borrow a phrase, the straw that broke the camel’s back.
The email was impeccably structured, confidently worded, and very obviously generated by ChatGPT. It had all the usual hallmarks: em dashes (“—“) liberally sprinkled throughout, selective phrases emboldened for emphasis, and that familiar, slightly formulaic tone AI produces when it’s trying to sound authoritative.
None of that, in itself, is a problem. I’m not anti-AI. I use it myself occasionally (even in writing this piece, ironically!). Where it crossed into unforgivable territory was the substance. The email cited legislation that (a) has never existed, (b) was superseded decades ago, and (c) only applies to England – despite the property being in Wales. A landlord without sufficient knowledge of housing laws could’ve easily been taken in by the e-mail.
“ChatGPT can make mistakes. Check important info”
The above message is displayed at the bottom of every screen when using ChatGPT: the most well-known and tested AI assistant, I’d wager. Unfortunately, in our experience, users are consistently ignoring that warning and treating its output as gospel.
And this isn’t just anecdotal. We are now seeing cases reach court where judges have explicitly noted that claims or defences appear to have been written wholly by AI. In some instances, these have been drafted by solicitors. In one case I read about recently, a landlord applied for a declaration of breach of lease. The leaseholder respondents cited three cases they believed were relevant to their defence – only they didn’t exist. The judge accepted this was not an attempt to mislead, but the lack of care went directly to the credibility of their evidence.
What’s striking is how quickly judicial tolerance appears to be diminishing. Early cases treated AI-generated errors as unfortunate but understandable. More recent decisions suggest a much firmer stance: if you put something before a court or tribunal, you are responsible for its accuracy – regardless of whether it was written by a human or a chatbot.
Against that backdrop, it should come as no surprise that we are seeing more and more tenants using AI when communicating with us, particularly where there is a dispute. Emails are longer, more assertive, and often framed in quasi-legal language that suggests the tenant is firmly in the right and the landlord very much in the wrong.
Where AI becomes dangerous
AI is ‘brilliant’ (cough!) at reinforcing conviction. If you already think you’re right, it will happily give you a well-worded explanation of why you’re right. If you’re doubtful, it will often convince you anyway. Where the underlying position is sound, that can be helpful. Where it isn’t, it creates a false sense of certainty that can escalate disputes unnecessarily.
From a landlord’s perspective, this trend is only going to grow. AI tools are now embedded in email clients, search engines and phones. Tenants don’t see themselves as “using AI” – they see themselves as getting help writing an email. That means landlords need to be alive to the fact that the tone and confidence of a message is no longer a reliable indicator of its legal accuracy.
What this means for landlords
What landlords cannot realistically – or legally – do is refuse to engage with correspondence simply because they suspect it was written by AI. Nor should they. The content still needs to be addressed, but it does need to be read carefully.
One of the biggest risks I see is landlords being wrong-footed by confidently written emails that, taken at face value, appear to set out a compelling legal argument. A well-structured paragraph with bold headings and statutory references can be surprisingly persuasive, even when it’s fundamentally wrong. Landlords need to retain their own conviction and avoid being swayed by presentation over substance.
Where this leaves landlords – and where CPS Homes adds value
A strong working knowledge of the law has never been more important. If tenants are going to arrive armed with AI-generated legal arguments, landlords need to be able to distinguish between what sounds right and what actually is right. That’s where professional support becomes critical. At CPS Homes, this is very much the space we operate in: correcting misunderstandings, pushing back on incorrect assertions, and making sure positions are grounded in the real legal framework, not a hallucinated one. All on behalf of our landlord clients.
Responses should be calm, measured and evidence-based. One of the downsides of AI-generated correspondence is that it can escalate tone very quickly. There is often an implied threat of action, complaint or litigation.
It’s also worth saying that this cuts both ways. Tenants are not the only ones using ChatGPT – landlords undoubtedly are too. We extend exactly the same word of caution to them. AI can be a useful research tool, a starting point, or a way of structuring thoughts. It is not a substitute for legal advice, nor is it something to rely on blindly. Anything sent to a tenant, tribunal or court should be checked, and checked again.
Closing thoughts
I suspect we’re heading towards a period where AI-generated communication becomes the norm, not the exception. The novelty will wear off, but the volume won’t. At the same time, scrutiny – from courts, tribunals and regulators – will continue to tighten. Carelessness, even if well-intentioned, will increasingly have consequences.
For landlords, the takeaway isn’t to fear AI or dismiss it outright. It’s to understand its limitations, recognise its influence, and make sure that decisions are driven by actual law and evidence rather than confidence dressed up as correctness.
As managing agents, that’s exactly where we add value. We act as a filter between AI-fuelled assertions and legal reality, ensuring that landlords are supported, informed and not unnecessarily pushed into conceding ground they were never required to give up in the first place.
In a sector already under pressure, that distinction matters more than ever.
Like what you’ve read? Keep up with our latest landlord tips and updates by signing up for our landlord newsletter.
The information contained within this article was correct at the date of publishing and is not guaranteed to remain correct in the present day.
