Universities must teach students to challenge AI, not defer to it
This digest was compiled by AI from multiple sources — links to the originals are below.
Generative AI is now used by 94% of UK undergraduates for assessed coursework, and more than half of 2025 scientific papers show linguistic traces of large language models. Researchers at the University of Oxford argue that universities must teach students to challenge AI systems rather than passively accept their outputs. The call comes as educators debate whether AI erodes critical thinking or can serve as an intellectual sparring partner.
Key Facts
- A March survey found that 94% of UK undergraduates use generative AI to complete assessed coursework.
- More than half of papers published in 2025 show linguistic traces of large language models, compared with about one in ten in 2023.
- Marc Schmitt, a researcher at the University of Oxford, argues that the danger is passive use of AI, not the technology itself.
- Times Higher Education reports that customised AI agents were designed to support active learning, reflective thinking and authentic engagement.
AI Adoption in Academia
In the United Kingdom, a March survey found that 94% of undergraduates use generative AI to complete assessed coursework. The scientific literature shows a similar shift: more than half of papers published in 2025 show linguistic traces of large language models, compared with about one in ten in 2023. This trend has alarmed educators and researchers, who worry that AI models could erode critical thinking and flood the scientific record with low-quality work.
The Case for Active Engagement
Marc Schmitt, a researcher at the University of Oxford who studies how AI reshapes decision-making in financial markets, argues that the alarmists are getting it wrong. Schmitt contends that the danger is not AI itself but passive use: treating the machine as a source of finished answers, when its value comes from active engagement. He notes that LLMs offer broad access to a capable counterpart that will reason with you on demand, redistributing intellectual firepower. Most researchers know that scientific ideas emerge through conversation and repeated attempts to articulate them clearly, and scientists can now have these intellectual exchanges with AI models at any time.
Pedagogically Designed AI Agents
Times Higher Education reports that rather than treating AI solely as a productivity tool, educators explored how customised AI agents could support active learning, reflective thinking and authentic engagement. Guided by the principle of 'pedagogy first, AI second', the team developed a series of pedagogically designed AI agents for undergraduate students. The agents were designed to encourage participation, reasoning and enquiry, rather than passively feeding students the answers. The design was informed by three principles: build with purpose, design for thinking, and keep it human.
2 sources
Time · lag behind first