Brits still aren’t comfortable handing over too much control to AI Christian Cawley | usagoldmines.com

• 84% of Brits fear handing too much control over to AI agents
• Concerns about the potential for AI agents to commit fraud or other illegal acts were also expressed
• 2,000 British adults were surveyed, with just 9% stating they have no worries about AI

The results are in and it’s clear: most Brits are uncomfortable with AI taking the reigns, and there are even concerns that an AI agent could turn to a life of crime – in your name.

Quizzed on their comfort using AI, 84% responded that they would give AI only moderate or low access to their lives, or none at all.

Instead, 27% of the 2,000 surveyed Brits are happier for AI to prepare things and wait for approval, while 23% are okay with letting AI carry out routine tasks, preferring to hold permission for the more important tasks instead.

Agentic fear

Exploring Brit’s concerns over the use of AI, the survey (conducted by Censuswide) found that 34% wouldn’t be comfortable with any AI control. But these results don’t mean that Brits are scared of AI. While they have “rogue AI” worries (40% believe the company that created the AI should be held responsible), only 19% consider AI to be a “threat to humanity.”

Meanwhile, just 9% say they’re not worried about AI at all.

Unsurprisingly, there is a split in demographics. In the over-55 group, 47% would refuse to give an AI agent control over their lives, while 20% of 25-34 year-olds consider themselves to be comfortable with high levels of AI agent control.

AI liability

But when it comes to AI potentially committing illegal acts, things look a little different, with 14% concerned about the impact of AI on their ability to stay out of legal trouble. It’s a figure that grows to 21% among the 18-24 year old age group.

“I think it’s important that we have a debate about AI liability – and this survey is important as part of that debate,” noted Jonathan Armstrong, a Partner at legal company Punter Southall. “Too many people assume that AI is the Wild West – legally that just isn’t the case.”

Consequently, it is people – those using the AI, who will be liable for criminal acts. “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.”

With the growth of AI agents offering a seemingly endless stream of automation possibilities, concerns from Brits should be no surprise. The onus is the user to understand what they’re using AI for, and to use it responsibly.