Students’ usage of AI reaches record high as academic confidence declines

The use of AI has risen exponentially amongst students, impacting academic confidence, critical thinking skills, and fluency. 

According to the Student Generative AI Survey (2026), carried out by the Higher Education Policy Institute, 95% of UK undergraduate students use some form of generative AI, such as ChatGPT or Copilot, and 94%  have used it for assessments. 12% of students use direct AI-generated text in an assessed submission in 2026, increasing from 3% in 2024.

Students face concerns over the work they produce … [as a chatbot] is seen as simply better

Agnieszka Piotrowska, Times Higher Education

The use of generative AI has been felt significantly within humanities and the social sciences, with these subjects having assignments that lend themselves to the functions of generative AI. However, its usage has been most common amongst STEM and healthcare students.

The use of AI in higher education has been a controversial topic, including concerns regarding the climate and ethics. However, often unspoken of is its effect on students’ academic mindset.

Students face concerns over the work they produce being incomparable to what an AI chatbot could, as the AI-produced work is seen as simply better. Whilst many tend to believe generative AI is a useful tool for saving time, others have begun to see it as a colleague, referring to a bot colloquially, and utilising it for even the simplest of questions.

The current grading system for higher education has proven a perfect match for generative AI, which can easily produce fluent, factually correct, and comprehensive answers – all highly desirable aspects of an assignment.

The problem is … a generation being rewarded for work that is AI generated to some extent, consequently creating the rhetoric that their own ideas are inadequate

It would be remiss to not also look toward the rise of AI within the professional workplace. An IBM study has found that 76% of organisations now have a chief AI officer, a significant increase from 26% the previous year, reflecting the scale of AI’s growth within professional workplaces. The stark changes do not go unnoticed. Is this uncertainty around academic output just a response to the current job market?

The way we use and perceive AI has changed, as is the function of AI within society, but does that have to be at the expense of human creativity and thought? 

Part of the problem is now grounded in a generation being rewarded for work that is AI generated to some extent, consequently creating the rhetoric that their own ideas are inadequate.

Detection of AI-generated content has become increasingly unreliable, and outright prohibition has been regarded as unreasonable, leaving institutions without a clear solution.