AI might soon be able to book our holidays, manage our diaries, shop for us and access our accounts, but new research suggests Brits aren’t ready to hand over the keys just yet.

A nationally representative survey of 2,000 UK adults, commissioned by Eskenzi PR and conducted by Censuswide, found that 84% would only give an AI agent moderate or low control over their day-to-day lives, or no control at all.

In fact, 34% wouldn’t be comfortable giving AI any control whatsoever.

Brits are much happier for AI to act as an assistant than take over completely.

Just over one in four (27%) would allow an AI agent to prepare things for them but would want to approve every action, while 23% would let it handle routine tasks independently but still require permission for anything more important.

Only around one in ten (11%) would be comfortable giving AI high or very high levels of control.

Could AI accidentally turn you into a criminal?

Brits have plenty of concerns about what could happen when an AI agent is given access to their accounts and allowed to act on their behalf.

The biggest fear is personal data being hacked or stolen, cited by 36%, followed by the AI simply making a mistake (31%) or having access to too much personal information (30%).

More than a quarter (28%) worry an AI agent could spend their money or buy something they didn’t intend, while 22% fear it could communicate with someone without their knowledge.

And perhaps most alarmingly, one in seven (14%) are concerned an AI agent could implicate them in something illegal. Among 18 to 24-year-olds, that rises to more than one in five (21%).

“I think it’s important that we have a debate about AI liability – and this survey is important as part of that debate,” said Jonathan Armstrong, Partner at Punter Southall.

“Too many people assume that AI is the Wild West – legally that just isn’t the case.  In most cases you’ll be responsible for the AI you use and the consequences – and sometimes even if it took you by surprise.

“There are many great use cases for AI but they should all start with knowledge of the risks and doing what you can to reduce that risk.” 

Older Brits tell AI: hands off

Age makes a big difference to how much control people are prepared to surrender.

Almost half of over-55s (47%) wouldn’t give an AI agent any control at all over their day-to-day lives. That compares with just 21% of 18 to 24-year-olds and 18% of 25 to 34-year-olds.

At the other end of the scale, around one in five 25 to 34-year-olds (20%) would be comfortable giving an AI agent high or very high levels of control. Among over-55s, just 5% feel the same way.

Would you ditch your bank for using AI?

The research also suggests businesses need to be careful about how they introduce AI agents.

Asked how they would feel if a company they rely on, such as their bank, insurer or utility provider, used an AI agent to advise its board on important decisions, 37% said they would want reassurance that a human always had the final say.

Almost a third (31%) would want to know exactly which decisions the AI had been involved in, while 24% would investigate how their personal data was being protected.

For some, it could even be enough to take their custom elsewhere. One in five (20%) would consider switching provider, while 18% would be less likely to buy additional products or services from the company.

By comparison, just 10% would see boardroom AI as a positive sign that the company was modern and embracing technology, and only 6% said it would make them more likely to use that company.

If AI goes rogue, who gets the blame?

The public also has a fairly clear idea of who should pick up the pieces if an AI agent is hacked, tricked or does something it shouldn’t.

40% believe the company that created the AI should be primarily responsible, while 25% point to the company or service the AI was accessing. Only 18% believe the individual using the AI should bear primary responsibility.

“AI agents promise to take a lot of boring jobs off our hands, and people clearly see the appeal of that. But there’s a huge difference between asking AI to organise your diary and letting it spend your money, access your accounts or make important decisions for you,” said Yvonne Eskenzi, co-founder of Eskenzi PR.

“What really stands out is that people aren’t rejecting AI. They’re setting boundaries. They want to know what it’s doing, they want someone to take responsibility when things go wrong and, when it really matters, they still want a human in charge.”

So, are Brits actually scared of AI?

Perhaps not in quite the way you might expect.

When asked about predictions that advanced AI could threaten humanity’s survival by 2030, the most common answer wasn’t fear of AI itself. More than one in five (22%) said they were more worried about what humans would do with AI than AI itself.

Around 19% believe AI should be treated as a genuine threat to humanity, while 18% think the danger is real but extinction by 2030 sounds far-fetched.  Just 9% said they weren’t worried about AI at all.

The findings suggest that, as AI agents become capable of doing more on our behalf, winning consumers over may depend less on showing them everything AI can do and more on reassuring them about what it can’t do without their permission.

The full report can be viewed, here.

Methodology: The research was conducted by Censuswide among a nationally representative sample of 2,000 UK adults between 16 and 18 September 2026. Censuswide is a member of the Market Research Society (MRS) and the British Polling Council (BPC), and a signatory of the Global Data Quality Pledge.