Letting agents will use AI, or be replaced by landlords who do. Their role isn’t vanishing, but evolving into something far more powerful. | AI Insights | Lumman.ai

Every industry has its “we’re not going anywhere” moment right before major disruption.

Travel agents had it. So did cab drivers. Remember when newspapers thought Craigslist wasn’t a threat?

Now it’s letting agents.

Letting agents tell themselves their work is safe from AI, because you can’t replace human relationships, right? The problem is — AI isn’t trying to replace them.

It’s giving landlords a reason to stop needing them.

Let's be honest: most lettings work is admin

Screening tenants. Chasing documents. Writing listings. Scheduling viewings. Negotiating offers. Following up on repairs. Logging maintenance requests. Sorting bank transfers.

It's all repetitive, rules-based stuff — perfect fodder for automation. Not someday. Now.

A mid-level landlord with 10 units used to spend hours on admin or pay 10–15% to a letting agent who largely did the same — just faster.

Now? That same landlord can plug into a productized AI that answers enquiries 24/7, runs financial checks, flags high-risk tenants, auto-suggests fair rents based on market comps, drafts the tenancy agreement, and even generates maintenance tickets from a WhatsApp message.

That’s not hypothetical. These tools already exist. Some landlords are already using them.

Here's what one said after his first three months using a GPT-based lettings stack:

“It’s like I hired a 24/7 back office assistant that doesn’t sleep, complain, or forget to CC the right documents. I've never had happier tenants — or less stress.”

So no — AI isn’t replacing letting agents. It’s making many of them optional.

The “relationship business” myth

Letting agents often argue their value is the human relationship.

But let’s unpack that for a second.

In most cases, the tenant barely meets the agent after check-in. And when they do, it's usually because something's gone wrong. Repairs. Noise complaints. Rent issues.

Those human-to-human moments matter — but they don't happen often. And unless the agent is really good, they’re not exactly delightful either.

Meanwhile, AI is getting weirdly good at mimicking polite, helpful, responsive customer service — at scale.

If you're a tenant with a leaking tap, would you rather:

Wait four hours for a human who may or may not reply

Get an immediate answer, an auto-scheduled plumber, and a confirmation text with ETA

Spoiler: Tenants don’t care whether it’s a human or not — they care whether their issue gets sorted. Fast.

That means the value of the agent can’t rest on "being a real person." It has to be something more.

The agents who win will look more like asset managers

Traditional letting agents did tasks.

The new generation will manage portfolios.

Think: instead of chasing lost keys and sending rent reminders, they’ll advise on yield, occupancy strategies, capital allocation, and real-time pricing. They’ll use AI systems not just to save time — but to find patterns across dozens of properties:

Which unit types are seeing the most churn, and why?
Which repairs are costing more over time and signal a future capex issue?
Which tenants are most likely to cause missed rent or disputes?
How should the portfolio be restructured to adapt to changing demand?

This isn’t property maintenance. It’s asset intelligence.

And ironically, it's only possible because AI has taken the grunt work off their plate.

A new kind of human edge

The real opportunity for letting agents isn't holding onto the past — it's moving up the value chain.

AI is freeing them from inbox management and scheduling hell. What they choose to do with that freedom will separate the replaceable from the indispensable.

Here’s how the best will use it:

Human strategy layered on machine precision
They’ll use AI to crunch the numbers, then add their own judgment on local market sentiment, regulatory nuance, or political risk. AI doesn’t know what a new borough policy might do to rental yields next year. Humans still win there.
Deep relationships that actually matter
Not check-in handshakes. Actual relationships with landlords, investors, and tenants — built on advice and problem-solving, not email threads. The kind of trust you call on when things go sideways.
Brand as a differentiator
AI is fast and cheap, but it’s not memorable. A strong brand, trustworthy communication, and a reputation for going the extra 10% suddenly become moats — the kind algorithms can’t replicate.

What happens next

Letting agents aren’t going extinct. But they also aren’t staying the same.

Their role isn’t being eaten by AI — it’s being redefined by it.

The agents who ask, “How can I use AI to do less admin?” will find small wins. The ones who ask, “How does AI change what letting agents should be?” will run the table.

Because here’s the uncomfortable truth: landlords don’t actually want an agent. They want outcomes.

They want high yield, low churn, fast payment, happy tenants, and peace of mind.

Who delivers that — a human, an AI, a hybrid — is increasingly irrelevant.

But the ones who try to hold onto every piece of the process manually? They won’t be let go because they’re bad. They’ll just quietly get replaced by someone faster, smarter, and cheaper.

And the landlord won’t even notice the difference.

Three things to think seriously about:

If AI can do 80% of your current role, what are you doing with the other 20%?
Make sure it’s something defensible — relationships, advisory, capital strategy — not just being available.
Don’t brand yourself by the tasks you do today.
Tasks get automated. Insights, trust, and judgment get prized.
The future of lettings won’t be about scale — it’ll be about leverage.
An agent with AI tools can manage 10x more properties with 10x more intelligence. That's not just efficiency. That’s a business model shift.

AI didn’t kill the travel agent. The internet did. AI just made sure people never looked back.

Don't be the agent everyone moved on from. Be the one they move toward.