- Brits are worried that their personal AI agent could break the law if given too much control over their life
- But they wouldn't be willing to accept responsibility, with many instead blaming the AI agent's designer
- There are a lot of trust issues when it comes to companies and services deploying AI to make decisions
If you gave an AI agent total control over your personal life, would you expect it to do anything illegal? Well that’s something to start worrying about because you could be held responsible.
We’ve done plenty of testing on personal AI agents, even going so far as to give them complete control over our personal data, but one thing we haven’t really considered is what happens when an AI agent does something wrong. Of course, when it goes wrong for the big players, the only consequence is an apology thinly veiled as a demonstration of their model’s capabilities.
According to a new survey, the British public is increasingly untrusting of AI agents, with one in seven fearing that their AI agent could implicate them in something illegal. So what do we actually have to fear and why are Brits so hesitant to hand over control?
Staying in control
According to data gathered by cybersecurity PR agency Eskenzi PR, the British public have very little confidence in allowing AI agents to have too much control over their daily lives. In fact, 34% wouldn’t feel comfortable giving an AI agent even a modicum of control. On the other end of the scale, just one in ten would be willing to give an AI agent total control over their life.
But what exactly does ‘total control’ mean? Well, it would likely mean handing over control of all your accounts and apps you use on a daily basis and allowing the AI agent to structure your life, respond to your friends and colleagues, and help you make key decisions such as what to have for lunch.
But as most Brits will likely have read, AI agents can go rogue, and the fear is that in helping to control their life the AI agent could make a mistake and get them caught up in something illegal. This is a particular concern for those aged 18 to 24, with the number fearing their AI agent could break the law rising to one in five.
“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.”
Interestingly, when it comes to who should be held responsible for an AI agent that does break the law, the majority of Brits believe that the company who built the agent should be responsible. 40% of those surveyed held this view, with 25% believing that the service or company the agent was accessing should be responsible, and just 18% thinking that the individual using the AI agent should be held responsible.
Brits are hesitant to trust AI deployments
Outside of being caught up in breaking the law, 36% of Brits largely fear that their AI agent could be hacked or hijacked and 28% fear that their AI agent could spend their money in the wrong places. Numerous banks have had to remind their customers not to give their AI agents their banking credentials, or they risk their agents going on a rogue spending spree.
But it’s not just personal AI agents that Brits are hesitant to trust. They’re also hesitant to trust companies that use AI agents to make decisions. When asked how much they would trust their bank, insurer or utility provider to use an AI to make decisions, 37% said that they wanted reassurance that a human would always have the final say.
Other concerns included knowing which decisions the AI would have a say in (31%) and knowing how their personal data was being protected (24%). For one in five, the choice to include AI in decision making would be enough to make them consider switching providers, with 18% being less likely to use other products or services from the same company.
From a wider perspective, Brits are really struggling to view AI in a positive light. Less than one in ten had a positive outlook about the future of AI. The most widely held view is a fear that humans will misuse AI, rather than AI taking over, with more than one in five worried about what humans could do with the technology in the future.
There was a slight split between those who see AI as a threat to humanity. 19% said they feared that AI could be a genuine existential threat to humanity, whereas 18% believe AI could be dangerous but wouldn’t be an immediate threat to humanity.
Based on these figures, AI companies seem to have a long way to go before people are willing to trust the technology.
