AI is already having seismic effects on Britain’s economy, but public and political opinion has yet to settle. This report draws on new research combining a nationally representative survey of 2,004 UK adults, a poll of 102 Members of Parliament, a dedicated survey of 502 UK technology professionals and a series of in-depth interviews with AI communications specialists.
Using a randomised control trial approach, we tested which arguments about AI actually shift attitudes, alongside analysing the concerns that shape public and political views. The result is a detailed picture of where the debate stands today – and what will be required to move it.
Executive Summary
The key findings from our research are set out below:
Britain is still persuadable on AI
Despite the confident views from the technology sector, public opinion on AI in the UK remains finely balanced. There is a winnable set of ‘swing voters’ but they are socially and economically distant from a thriving, young, educated technology sector. This makes the outcome of the debate contingent – not inevitable – and highly sensitive to how AI is communicated.
Safety – not jobs – is now the focus
While job displacement is often assumed to be the central anxiety around AI, it is far from being the most pressing worry today. Instead, concerns about safety, misuse, fraud and loss of control loom larger for both politicians and the public, including those who have adopted AI in their daily lives.
Only one argument really changed minds
When tested, only one of the pro-AI arguments explored produced a measurable shift in public opinion. It increased support for AI from a ‘referendum-losing’ 45% to a ‘referendum-winning’ 56%. This effect was observed across partisan and demographic groups. The same argument also emerged as the most persuasive among Members of Parliament.
Winning on AI means winning in the NHS
This winning argument focussed on AI’s potential to revolutionise public services. Crucially, this is not about abstract transformation, but practical outcomes – shorter waiting times, greater efficiency and better frontline delivery in the NHS. Across age groups and political divides, this relatively modest promise proved more powerful than more dramatic visions of how AI might transform society.
The tech sector is misreading its core audiences
Across multiple measures, technology professionals struggle to accurately identify what most concerns and persuades both the public and politicians. Fewer than 2% of technology professionals polled could identify more than half of the most compelling arguments with the public or politicians – and the more senior they got, the more they struggled.
Persuasion, not just innovation, will determine AI’s future
Advances in capability alone will not secure public or political backing for AI. Support will be shaped by how effectively the benefits are translated into outcomes people understand, trust and value. Organisations that pair innovation with compelling, resonant communication will be best positioned to build lasting consent for AI adoption.