The challenge
Like many council communications departments across the country, the small team at West Oxfordshire District Council is tasked with serving an ambitious administration and looking to deliver the best possible communications to residents, while keeping demands on the public purse as light as possible.
The core comms team – composed of three people working alongside shared resources – had used some elements of AI beforehand as individuals but never looked at it strategically and across all their work. They also didn’t have a proactive organisational approach to AI, so they decided to look at comms as a specific service. This included reviewing all of their processes and outputs to see where AI could add value.
The team is open-minded about AI, as part of a broader approach of exploring every opportunity to improve comms with residents and ultimately deliver a better service. Following a presentation about the potential for AI at a Local Government Association event in September 2024, they decided to look into it in more detail.
The solution
Early tests had shown potential for value-add but also limitations. They were also conscious of the governance, data security and other aspects of AI use that would need to be addressed through the work. Areas were identified where AI could add value without introducing any significant risks, and help to free up time for the team to focus more on strategy, evaluation and creative work. These include writing draft press releases, copy for newsletters and social media, producing comms plans, and help with writing speeches and presentations.
The team spent around a week in total, spread over several months, analysing what their requirements might be, choosing an AI platform, learning how to use it and developing an approach. They decided on ChatGPT, at a cost of about £1k per year.
The results
The time savings that have been achieved through the use of AI add up to the equivalent of up to an additional member of staff, the team estimates.
Mark Pritchard, head of comms and corporate strategy at West Oxfordshire District Council, has been delighted with what has been achieved. “AI gives us the ability to deliver our work more efficiently,” he explains, “enabling the team to focus more on the creative and strategic parts of their jobs.”
He adds: “For a small investment in time and funding, we are getting a worthwhile benefit for all of our team members that will ultimately help us deliver better comms to our residents. It is helping our small team do more and focus on the strategic and stakeholder work that AI can’t support effectively.”
Taking advantage of what AI can do is a continuous process, and the team is looking at other ways that AI can be used to support delivery such as in design and day-to-day administrative tasks.
“AI is improving so quickly it's hard to judge where it will be in a month, let alone [in the] longer term,” says Pritchard. “I am certainly no expert – just someone trying to keep up and make best use of this exciting new technology.”
Caution required
While the benefits have been tangible, AI is certainly not a panacea, the comms team has found. The real benefit lies in repeatable tasks, rather than one-off scenarios, where it can take as long (or longer) to design a prompt that will give a good-quality output as it would take to do the job yourself.
Furthermore, the team says, having a well-thought-through AI policy is crucial to ensuring the risks of AI are managed. It is also vital that AI-produced content is proof-read thoroughly, as sometimes the work it produces is of poor quality and occasionally includes something that wasn’t in the source material.
The team has learned that the quality of the output entirely depends on the quality of the prompt and the reference material the AI is given, and that it needs to be given both good and bad examples to help it understand the tone and style it needs to replicate.
While AI can be “incredibly helpful, it still has significant limitations”, Pritchard warns. These include making mistakes that can be easily missed if not carefully monitored. Without proper scrutiny and attention the new technology also “poses risks to data protection, can lack the context needed to give insightful responses, and can at times raise ethical concerns surrounding trust and authenticity”, he adds.
Rather than simply being the threat to jobs many people imagine, AI can actually make people more efficient, according to Pritchard. However, he stresses: “Human oversight remains crucial for setting instructions, guiding its use, and monitoring its outputs."
Photo: Getty Images