AI Isn’t the Problem. Over-Reliance Is.

“We will get to a point where we will love it when we see spelling mistakes.’

Prophetic words that were spoken to me around 18 months ago by one of the UK’s best agents.

She was speaking about the rise of AI.

Fast forward to this week (17 March 2026), I read an article about the movement/race to create a universal brand, highlighting where AI is (or, more importantly, isn’t) used to create something, be that a film, a song, or even an estate agency-related blog.

This is something I suggested back in March 2023 when we launched the EAN – we even designed a logo for it – ‘Written by humans’.

It feels like chickens are coming home to roost.

There’s no denying AI has had a massive effect on our business.

We closed the Estate Agent Content Club down last month, as I quickly realised that you can’t build a business solely around content.

It’s why the Ethical Agent Network is growing – built on a mixture of four C’s:

Credibility, you can’t just buy your way into the EAN. There’s a genuine application process.

Courses – education by agents for agents is a big part of what we’re doing.

Community – peer-to-peer support, private WhatsApp groups and online and in-person get-togethers.

Content – yes, content still plays a part in our model through blogs and guides.

The problem isn’t AI, it’s the growing over-reliance on it – for the record, we use AI for several things within the EAN.

Our EAN-based articles in our weekly content packs are always written by a human (his name is Mark; he’s good), but for full transparency, I check and edit them using a combination of my editorial experience and Grammarly (which is AI, ultimately).

Our bonus articles on breaking news, such as interest rate decisions and budgets, are written by humans and always will be.

Budgets are tight at the moment in the EAN as we’re in the growing phase, and AI is helping carry a fair chunk of our content load. But when we hit 50 members (we’re at 40 as of 17 March 2026), we will revert to all our articles being 80 per cent written by humans – the other 20 per cent is the proofing/editing that AI excels at.

Since AI’s arrival with ChatGPT, we have always tried to use a Human Sandwich when approaching content creation.

Firstly, we feed some ideas into ChatGPT (we’re also looking into Claude, as we’re hearing good reports about it and it seems more ethical than its rivals). We then let ChatGPT generate ideas, some good, some sh!7e. Off that back of that we will ask it to draft articles in our tone of voice and style – that’s the sandwich filler before another layer of humanity is applied when we go through it and edit it (using a mix of our knowledge and Grammarly spell checker for transparency).

We also use AI to refresh our Daily Ideas, since we have more than 400 in the bank and they often need a little glow-up before reposting to remind people. Interestingly, non-AI-generated ideas seem to work far better in terms of engagement.

Here’s the number one way that agents can still produce good content without resorting to AI slop.

Create more video content.

Such as:

• To camera videos – yes, you’ll, um, err and have outtakes, but it shows you’re human.
• Interviewing local businesses.
• Talking about community projects.
• Creating behind-the-scenes content – for example, drawing back the curtain of what happens in your agency.
• Reaction to camera pieces – it doesn’t have to be something that happened nationally – locally is better.

Practising what we preach, we’re setting up a small filming ‘studio’ in our office to make it easy for me to record videos.

 I say a studio, it’s hardly Warner Bros, it’s just a good mic (dji Mic 3), a decent right light (NEEWER ring light tripod) and my iPhone 17. Watch this space (literally).

AI isn’t the issue – it’s laziness. We’re all guilty of it, but to stand out in a world drowning in AI slop, we need to put a little more effort and creativity in, my company and I very much included in that.

Thanks for reading.

Jerry

PS – The process for this article was me spilling my thoughts into Word, then putting the rough draft into ChatGPT. With this prompt, “Give me 10 non-AI sounding headline options, tighten up the copy to 600 words maximum, correct any grammatical errors.  And avoid any wanky AI phraseology.”

 What it came back with, annoyingly, wasn’t what I wanted, so the blog above is my warts-and-all (Grammarly-checked) version.

Here’s the BBC article: https://www.bbc.co.uk/news/articles/cj0d6el50ppo