Younger people are often assumed to embrace new technology with enthusiasm. Yet the return of wired headphones and the wider ‘analogue renaissance’ suggest that might no longer be a given. Public First’s AI Global 2026 report found that AI is no exception: net positivity among British 18-24 year olds has fallen from 63% in 2023 to 46% in 2026.
What makes this surprising is how much they still use it. Almost two-thirds (63%) of 18-24 year olds use generative AI at least weekly, including nearly one in three (32%) who use it every day.
So what explains this? Familiarity has not translated into optimism, nor are young people struggling to see AI’s immediate value. More than half (53%) continue to use it because it saves time. The tension seems to emerge when they look further ahead.
Across our polling we found some common themes:
Young people are worried about what AI means for gaining workplace skills. More than a third (36%) of 18-24s worry AI will make critical thinking harder, while 33% think developing expertise will become more difficult. Overall, 60% of young adults worry at least sometimes about relying too much on AI.
Career stage appears to shape views of AI. 18-24s and 35-44s use AI at almost identical rates, yet only 19% of the younger group believe it will improve both their own opportunities and society, compared with 28% of 35-44s.
AI optimism follows economic optimism. 46% of respondents across all age groups who expect their personal economic situation to improve ‘a lot’ also take the most optimistic view that AI will benefit both themselves and society, compared with only 16% amongst those who expect their economic citation to get ‘a lot worse’.
Familiarity Does Not Always Produce Optimism
Previous Public First research, found that targeted training and workplace experience can close demographic gaps in use, such as amongst older women, and that this in turn improved optimism about the role of technology in society.
But 18-24s are already among the most frequent users of generative AI. The comparison with 35-44s makes this particularly striking. Despite almost identical levels of use (63% vs. 61% say they use generative AI tools at least weekly), younger adults are far less likely to believe AI will improve both society and their own opportunities (19% vs. 28%). Similar exposure is producing different expectations.
Career stage offers one explanation. By their mid-thirties and forties, many workers have expertise that exists independently of AI, allowing the technology to complement capabilities they already possess. Younger workers are encountering the same tools while trying to build that foundation.
The Value of the Entry Level Job
Generative AI is particularly good at many of the tasks associated with early career: summarising documents, drafting emails and turning notes into a report. An experienced worker already has the judgement to assess AI’s output and fill the gaps. The technology can speed up work without necessarily replacing the process through which professional judgement develops.
For someone at the start of their career, the same shortcut is more ambiguous. AI can improve today’s output while reducing the opportunities through which tomorrow’s expertise is built.
Our polling suggests young adults are thinking carefully about this. More than a third (36%) worry AI will make critical thinking harder, while 33% think it will make developing expertise more difficult. Overall, 60% of 18-24s say they often or sometimes worry about relying too much on AI, compared with only 39% of 55-64 year olds.
This matters because entry-level work has always done more than produce entry-level output. It is the opportunity for young people to learn by doing, have their work reviewed by more experienced colleagues and gradually develop their own judgement. Even where junior jobs remain, what those jobs teach will change if new entrants are expected to produce work at a pace that only AI can facilitate.
A Participation Paradox
None of this is leading young people to opt out of using AI. More than a quarter (30%) have used AI for job applications or CVs and 25% for workplace tasks.
But opting out increasingly carries a cost. More than half (52%) of 18-24 year olds have felt pressured to use AI, higher than any other age group.
This is the participation paradox at the centre of young people’s relationship with AI. The technology is useful and increasingly expected, yet it has arrived at a point in their careers where they want to develop the skills that managers are tempted to use AI to shortcut.
That tension may also help explain why the majority view amongst 18–24s is expecting AI to improve their own opportunities while harming society (35%). They can see what the tools do for them today, while questioning what greater reliance might mean for the workforce that follows.
Opportunity Shapes Optimism
There is a wider economic context to this anxiety. Across all age groups, 46% of people who expect their personal economic situation to get ‘a lot better’ also expect AI to improve both their own opportunities and society. Among those who expect their situation to get ‘a lot worse’, just 16% say the same, while 58% expect AI to harm both.
Current financial circumstances tell a similar story. Among those who describe themselves as very financially comfortable, 43% expect AI to improve both their own opportunities and society, compared with just 18% of those who regularly go without essentials such as food or heating. The pattern also travels across borders. Public First’s AI Global 2026 report found higher optimism in faster-growing emerging economies than in slower, established economies, typically in the West. Where people already expect growth to improve their lives, technological change is easier to see as an opportunity.
This does not prove that insecurity causes pessimism. But it does show that views on AI are being formed through people’s existing experience of the economy. British young adults are entering this transition before accumulating many of the financial or professional buffers older workers may have. Their concerns about AI therefore often arrive on top of existing uncertainty.
Good adoption vs bad adoption
The debate about AI has largely focused on adoption, how to get more people using it, more often. The harder question is about the quality of this adoption.
The economic case for AI is significant. Public First’s previous research found that, all else being equal, people who use AI effectively achieve better outcomes. Realising those gains depends not just on widespread adoption, but on ensuring people have the judgement and expertise to use these tools well.
Our polling shows that many young people are not yet convinced the current transition is creating those conditions.
Some of the pessimism also reflects a broader insecurity about economic prospects that AI policy alone cannot solve. The more immediate challenge is the UK’s talent pipeline. Employers need to think carefully about how AI enters the workplace, preserving space for junior workers to build expertise and therefore thinking carefully about how this sits alongside the rollout of AI.
If the generation entering the workforce comes to associate AI with pressure and fewer chances to learn, Britain risks accelerating AI adoption in the short term, but weakening the talent and expertise on which its long-term economic benefits depend.