Survey: a third of university students in Great Britain predict civil unrest from AI job losses
Fears of job losses and unrest
One in three university students believes that artificial intelligence will eliminate jobs so rapidly that it will trigger civil unrest, according to a new survey by King’s College London (KCL). The figure, 34%, is significantly higher than the 22% of the general public who share the same concern. More than half of the students surveyed said they were convinced that job losses caused by AI would be worse than in a normal recession, with the KCL report noting that this pessimism reflects a belief that AI will continue improving and taking roles faster than workers can recover.
The threat is felt most acutely in entry-level positions. According to the KCL survey, nearly six in ten respondents across all groups — university students, young people, employers and the public — agree with the prediction that AI could eliminate half of all entry-level white-collar jobs within five years. Among university students, that proportion rises to 63%, highlighting the extent to which those about to enter the workforce expect the ground to shift beneath them.
The findings come from the first wave of a new tracker of attitudes to AI, run jointly by the King’s Institute for Artificial Intelligence and the KCL Policy Institute. It polled 1,000 university students, 1,000 young people aged 16 to 29 in England, Wales and Scotland, 500 employers, and 2,000 members of the public across the UK.
Heavy usage and common pitfalls
Students are among the most intensive users of AI tools. The survey found that 77% use AI at least a few times a month, compared with 46% of workers, and 27% use it daily or almost every day. Yet their familiarity with the technology has not shielded them from its shortcomings. Nine out of ten students said they had encountered problems when using AI — most commonly factual errors or inaccuracies (37%) and fabricated sources or statistics (31%). Despite these widespread issues, fewer than half of students said they usually or always check AI output before using it in their work, raising concerns about the quality of the material they may be submitting or relying on.
A notable gender divide exists in students’ perceptions of AI. While 52% of male university students said they believed AI was a positive thing for humanity, only 39% of female students agreed. Male students were also the most confident of any group polled that AI was improving their ability to think for themselves; female students were the most likely to feel the opposite.
Optimism and preparedness
Despite their deep anxieties about AI’s impact on jobs, students remain broadly committed to higher education. The survey found that 78% of those who took part would still choose to go to university if they were starting again, though 30% said they would have picked a different subject and 12% said they would not attend at all. There is a clear gap, however, between what students think universities can do and what they are actually delivering. While 60% said they believed universities were capable of preparing them for an “AI-shaped job market”, only 36% said they were receiving that preparation. This disconnect is mirrored at the national level: only 20% of the general public think the education system is adequately equipping young people for an AI-influenced world, according to broader polling cited in the KCL research.
The UK government has responded with targeted investment, including £4 million to develop AI tools that reduce teacher workload and £1 million for innovative data applications to support feedback and marking. The Department for Education has also introduced an “AI Product Safety Expectations in Education” framework to ensure tools are secure and ethical. Universities, meanwhile, are adapting curricula by introducing interdisciplinary modules on AI ethics, data literacy and industry-focused AI skills, though the KCL study notes that institutional policies remain fragmented, with some universities offering robust guidance and others providing minimal direction.
Bobby Duffy, director of the KCL Policy Institute, said the findings show that “the public, workers, young people and university students are watching the rapid development of AI with more fear than excitement, with real concern for what it will do to jobs, particularly at entry levels, and, therefore, the prospects for our young people and the economy in general.” He added that governments and employers must now demonstrate how the labour market can adapt to AI disruption.
Bouke Klein Teeselink, a lecturer in philosophy, politics and economics at KCL, struck a more hopeful note. “With the right training, policies and institutional support, there is a clear path forward to a more hopeful future, with rising productivity, broader opportunity, higher incomes and faster scientific progress,” he said.
