The survey of 2,000 people aged 18 and over also found that 75% of respondents want to be told when they are not ‘dealing with a real person’ and 67% think that AI should not ‘pretend to be human or act as if it has a personality’.

The majority of people surveyed also agreed that automated AI-driven campaigns should be ‘carefully regulated’ ( 72%) and 61% agreed that people must accept liability if the use of AI results in an accident.

Half of people surveyed ( 51%) agreed that AI should have the right to report them if they are engaging in an illegal activity, the survey found.

Comparing the results with a 2018 survey conducted by Opinium and the IPA with the same question, there has been a 25% decrease in the number of people who think that they should be polite and exhibit good manners when interacting with virtual assistants, from 64% in 2018 to 48% in 2023.

Josh Krichefski, IPA president and chief executive, EMEA & UK, GroupM, said: “AI provides incredible opportunities for our business. As these findings demonstrate, however, the public are understandably cautious about its use – and increasingly so in some areas.

“It is therefore our responsibility to be transparent and accountable when using AI to ensure the trust of our customers.”