Last week, the Department for Science, Innovation and Technology revealed that its artificial intelligence tool could free up around 75,000 civil servant days every year. It’s a striking demonstration of how AI can deliver measurable efficiency gains across government – if it’s implemented in the right way.
But while the figures are impressive, they only tell part of the story. True transformation won’t come from automation alone; it will depend on how successfully the civil service supports people to adapt, build confidence and make the most of these technologies in their day-to-day work.
Because the real challenge facing government isn’t just to deploy new technology; it’s to build a culture where AI strengthens, rather than sidelines, the human expertise that underpins great public service.
Across thousands of workplaces in the UK, AI is already reshaping day-to-day work, particularly in high-contact environments like customer service. In recent months, we at ArvatoConnect undertook a study – with the findings detailed in our Impact of AI on Agents report – into how frontline service professionals are experiencing this shift. While not exclusive to government, the findings offer timely lessons for those leading transformation across the public sector.
ArvatoConnect’s recent work with the Crown Commercial Service shows how embedding AI alongside human expertise can improve accuracy and free people to focus on higher-value tasks.
What’s clear from our research is that AI is already embedded in day-to-day operations, with 94% of respondents using AI tools as part of their roles. Many say the technology is making their jobs easier. Around 70% said AI helps them resolve complex issues, while nearly two-thirds believe it improves the customer experience. More strikingly, 67% said their job satisfaction had improved as a result of using it.
But the research also revealed some important challenges. A significant number of workers felt uncertain about their future – not because they resisted the technology, but because they didn’t feel they had been brought along with it. Patchy training, unclear communication and a lack of visibility on career development were recurring themes, with one in four saying they had considered leaving their current role because of AI-related concerns. A further 30% said that although they weren’t planning to leave, they were uneasy about the changes taking place.
It is easy, particularly under pressure, to modernise too quickly, prioritising technical rollouts ahead of the cultural work that must be done in parallel. Doing so risks leaving the workforce behind. The key to sustainable success is ensuring AI integrates into organisational culture, rather than defines it.
Transformation isn’t just about software or technological advancement; it never has been. It’s about how people adapt, how they are supported through change and how transparently and clearly the benefits are communicated.
We need to recognise that AI brings more than productivity; it brings change in workflows, expectations and the skills required to do the job. If that change feels opaque or unbalanced, people disengage. For the civil service, this is especially important. These are services delivered in sensitive and often high-stakes contexts, so ensuring that employees understand, trust and feel equipped to use AI is not a “nice to have”; it’s foundational to success.
The solution lies not in slowing down, but in being more considered in the use of AI. Embedded training must evolve with technology. This means covering the how, why, what, when and where of AI, and being honest about how roles will adapt – involving workers early, listening to concerns and showing where they fit in the future. The people delivering these services every day understand where the real pain points lie, and if involved from the outset, can help stress-test ideas and shape AI systems that are practical, usable and effective.
There is no shortage of optimism about what AI can deliver. But optimism alone won’t sustain a workforce. The next phase of government transformation must focus as much on people as on process, because only when employees are engaged, supported and confident will the full promise of AI be realised.
If we want to deliver truly intelligent public services, the implementation of AI must involve people – not as an afterthought, but as a guiding principle.