- New polling from the Institute and Ipsos finds that
39% of Britons see AI as a risk to the economy, while 20% see
it as an opportunity.
- Polling finds strong correlation between AI usage and belief
in the technology's potential, with 56% of those who have never
used the technology seeing it as a risk, compared to just 26% of
those who use it weekly.
- Critically, 38% cite a lack of trust in AI outputs as the
main barrier to adoption, higher than any other factor.
- User patterns vary widely. 25% of UK adults use generative AI
at least weekly. However, nearly half of UK adults have not used
any generative AI tools at all in the past 12 months
- With AI central to the government's growth agenda, the
Institute argues that it must build public confidence with a
high-priority national programme of AI education and outreach to
‘clear the fog' surrounding development.
Almost twice as many Britons see AI as a risk to the economy than
an opportunity, according to new polling from the Institute (TBI).
In new paper, ‘Building Public Trust in AI to Accelerate
Adoption', the former Prime Minister's Institute warns that
without decisive action to include the public in AI development,
a lack of trust will undermine the government's growth agenda and
leave disadvantaged groups behind.
The polling by TBI and Ipsos reveals that 20% of Britons see AI
as an opportunity for the economy, while 39% see it as a risk.
Critically, it finds that trust and optimism about AI are closely
correlated with regular usage, which in turn is linked to being
younger, male and higher income.
TBI's paper argues that with AI adoption central to the
government's growth strategy, as set out in the AI Opportunities
Action Plan, ministers must take urgent steps to ensure the
entire population is equipped with the confidence and skills to
benefit. Otherwise, public scepticism will undermine progress.
The survey of 3,727 UK adults reveals that AI usage in the UK is
highly varied; while more than half of respondents had not used
AI in the last 12 months, 26% use it weekly. Critically, 38% cite
a lack of trust in AI outputs as the main barrier to adoption,
higher than any other factor.
The polling demonstrates that usage and confidence are sharply
divided by age, gender and income. Adults under 35 are far more
likely to view AI positively, while older demographics are
significantly more sceptical. Women are six percentage points
less likely than men to see AI as an opportunity for society,
even when accounting for income and education.
Confidence in using the technology is also associated with
income, with those higher earning most likely to say they're
confident in AI, even in sectors most exposed to disruption such
as finance, professional services and IT.
The findings highlight a striking correlation. 56% of people who
have never used AI see it as a societal risk, compared to just
26% of those using it weekly.
This ‘experience gap' means that those already benefiting from AI
are more likely to trust it, while others remain sceptical.
Jakob Mökander, Director of Science and Technology Policy at TBI,
said:
“We simply can't ignore the opportunity AI presents for growth –
in building a thriving tech sector, attracting more capital to
the UK, and enhancing productivity across the economy. Off the
back of last week's deal with the US, Britain has the perfect
window to show the world what's possible, and lead in tech
diffusion and application.
“To realise this, we need to prove to people that AI will work
for them, not happen to them. We need to clear the fog and
mystique around AI development and show people it's a helpful
tool just like any other.
“Building this trust across the entire country, not just those
who are already evangelists for AI, must be an urgent priority if
we want to bring the benefits of AI to Britain, and realise the
government's growth agenda.”
To bridge the trust gap, the Institute argues, the government
must follow a strategy of active public engagement combined with
policies that build justified trust in AI.
The authors recommend a two-pronged strategy. First, the
government must ensure that AI systems are worthy of the public's
trust. This means ensuring that all AI tech is robustly tested
for safety, with proportionate regulation in place to protect
people's interests.
Second, the government must deliver a national programme of
outreach and communication efforts to improve public attitudes.
At the centre of this strategy should be the principle of talking
about AI in a ‘human' way; communication should not get stuck on
infrastructure and the ‘nuts and bolts' of the tech but showcase
how it can help in the day-to-day.
Also recommended is a national rollout of AI training programmes
developed with employers, unions, professional bodies and further
education providers that target people of all backgrounds, not
just those already working with AI. Efforts should also be
supported by engagement initiatives, potentially including ‘AI
Open Houses' that invite citizens into cutting edge AI labs, and
an accessible, publicly broadcast lecture series.
Mökander concluded:
“There is currently a gap between the government's narrative and
public attitudes towards AI. Building trust in AI will be key if
the government is to succeed in delivering the AI Opportunities
Action Plan and its wider growth agenda.
“Most people don't need or want to know the technicality of large
language models, semiconductors and neural networks. If we talk
about AI only at this level, we just put up more of a
smokescreen.
“What they want is reassurance that the technology is safe to use
and works in their interests. They want to understand how it's
going to improve their lives; their children's education and
career prospects, their jobs, their pockets, and the way they
interact with government.
“Governments have a legitimate role in not only ensuring that AI
is safe and effective, but also in building healthy public
attitudes towards the risk and benefits. With good governance and
increased public engagement, we can build the trust needed for
the government's growth agenda.”
--ENDS---
Notes to editors
- The polling was conducted by Ipsos Knowledge Panel on behalf
of TBI between 30 May and 4 June 2025, with a nationally
representative sample of 3,727 UK adults.