By Samantha Bennett, Media and Communications Specialist, Wythenshawe Community Housing Group

Twelve months ago, our communications team sat in an audience at a conference listening to a talk on the so-called Fifth Industrial Revolution. No pressure, then. The speaker, Dan Sodergren, made AI sound both terrifying and irresistible. Listening to Dan, I knew immediately this was going to change how we worked, both individually and as a team and organisation.

If you’re eyeing up AI tools but unsure where to start, here’s what we’ve learned after a year of using them almost every day.

First, the human in the loop isn’t going anywhere. I assumed AI would leave carefully researched facts alone. It doesn’t. It happily rewrites dates, invents statistics, or swaps names without blinking. A beautifully phrased line is no good if it’s completely wrong. So yes, AI saves time, but it also needs proper oversight.

We started small, tested carefully, and invested in a secure organisational account so our business and residents’ information weren’t feeding public models. From the start, we’ve anchored to our AI policy, making sure we use these tools responsibly and ethically. Early buy-in from our executive director of transformation and resources and the ICT team made rolling out Chat GPT to the team easier and gave us full support.

We didn’t throw AI at everything. Instead, we focused on where it could deliver quick, practical improvements. Drafting and sorting content is relentless, internal weekly emails, intranet and website updates, resident newsletters, and safety leaflets. AI has helped us claw back hours each week. When you’re a team member down, as we have been, that’s more than a nice-to-have. It’s been essential. We also tied AI to our tone of voice project, This Is Us, to keep what we wrote clear, genuine, and human.

One lesson stands out: considered beats scattergun. We’ve now built and tailored over 20 reusable prompts, our little team of AI interns working behind the scenes. They’re all tested and refined so everyone in the team can trust them. They don’t replace judgement. They free us to use it where it counts.

Despite the wins, using AI for us is still a work in progress. Most of our use is content creation alongside task management that helps shape clear, prioritised plans. Automation is next on our to-do list, finding better ways to streamline all the copying and pasting between systems. And we’re staying clear-eyed about the limits. AI loves to flatter, waffle, and even ignore instructions. Think of it as a very enthusiastic intern, keen to help but prone to people-pleasing, not to mention forgetting what you asked for or why it matters. Oh, and it loves to add its own flair with creative embellishments.

It’s also worth remembering the environmental impact. Every time we fire up an AI tool, it draws on huge data centres, consuming water and power most people never see. It’s not free, financially or environmentally, and in a sector that cares about sustainability and reducing our carbon footprint, that matters. Just because you can generate five drafts in 30 seconds doesn’t mean you should.

Looking ahead, there’s loads we haven’t explored yet, but for now, we’re focusing on embedding what works and staying curious about what might be possible.

If you asked me whether you should bother trying AI, I’d say absolutely. Explore, play, start small, and pay for a secure workspace. Accept that you’ll need to review everything and that it’s not the answer to all. And remember, no matter how clever the tech gets, the human in the loop is staying put.

That’s probably the biggest thing I’ve learned, AI doesn’t replace communications professionals. It respects their time. It gives them space to think, create, and add value, the stuff we all came into this sector to do.

Just don’t expect it to get your dates right. Some things never change.

I’d love to hear from others in the sector who are trying new things with AI or exploring different approaches, let’s swap notes.