Minority of UK consumers still use AI daily

Even after several years of hype, the reputation of artificial intelligence as the technology of the future does not seem to be translating into actual use among consumers. Even amongst the most enthusiastic of demographics, only about a quarter of people are using the technology in any form on a daily basis.

For the past two years, AI hype has come with a caveat: to unlock the potential of the technology, more capacity will have to be opened up to power AI solutions. The UK’s Labour Party, which was voted into government in the summer, bought into this wholesale, and has recently re-confirmed its commitment to give new data centres priority – pushing back regulatory measures to enable data centre projects to build on green belt land via a fast-tracked scheme that would prevent local councils from blocking new sites. Meanwhile, the emergence of DeepSeek as a cheap alternative to ChatGPT means the technology has been at the top of the news again over the last two weeks.

Even so, recent surveys suggest the proliferation of the technology into public consciousness is not necessarily leading to more usage. In 2024, a Deloitte poll found that show that beyond brief dalliances with the technology, the vast majority of people in Britain do not find it useful in their daily lives – and 74% of respondents said they had not used generative AI in particular. Now, a new poll of 2,003 people across the country has found that little has changed heading into the new year.

Interviewing more than 2,000 UK consumers, the research from TalkTalk found that half of respondents did not recognise the term ‘AI’ as familiar – and just 18% were using it daily. Lower still, only 15% said they used AI daily – and even when breaking the results into demographic chunks, a majority never said they used the technology regularly.

Familiarity rose among those aged 16 to 44 years old, predictably higher than the 32% of those over 55. But this familiarity is not translating into regular use according to TalkTalk’s findings, with “the highest AI users” being those aged between 25 and 34 – of whom only 24% were using it every day.

Of the respondents using AI for tasks of any kind, people seemed most willing to use it as a search engine. 29% of respondents said they deployed AI such as ChatGPT to help them with online shopping recommendations. With the degeneration of leading search engines such as Google – where algorithms prioritise keeping users on the platform for longer, rather than deploy the most helpful information – that may be the biggest opportunity open to AI vendors; though 29% is still a long way off the levels of use needed to return on the massive investments needed to create the technology in the first place.

At the same time, the technology’s prized generative functions still seem to lag behind in popularity. As impressive as they have been said to be, only 28% of consumers would trust an AI to draft a post for social media – while an even slimmer portion of 26% would use it to compose work emails; a front on which a great deal of hope for AI’s potential profitability has always hinged. But its propensity for ‘hallucinations’ (mistakes) mean that in the most important cases, it seems that there is still very little inclination to deploy it in this arena.

This may be about to change, with the advent of DeepSeek, however. As the technology is open source, the previously opaque world of GenAI may be about to open up – and that could be key to build the kind of trust needed for greater engagement. Trust in AI currently stands at 45%, while 36% express neutrality and 17% distrust its reliability. This significant neutral segment underscores the importance of building public confidence in AI.

TalkTalk’s study notes to that end, “Transparent AI systems that explain their operations can help transform undecided users into advocates. Additionally, gender and regional differences highlight gaps in outreach and education that must be addressed to foster inclusive trust in AI technologies.”