British universities have been advised to perform “stress-tests” on all assessments as new research has shown that the majority of undergraduates are utilizing generative artificial intelligence (genAI) in their studies.
A study of 1,000 students, both domestic and international, found a significant increase in the use of genAI in the past year. Almost 9 out of 10 students (88%) in the 2025 survey admitted to using tools such as ChatGPT for their assessments, compared to only 53% last year.
The use of any AI tool has also risen from 66% in 2024 to 92% in 2025, meaning only a small percentage of students are not using AI, as per a report by the Higher Education Policy Institute and digital e textbook provider Kortext.
Josh Freeman, the author of the report, stated that such significant changes in behavior within a short span of time are uncommon. He emphasized that universities need to acknowledge that genAI is here to stay.
Freeman highlighted the urgent need for institutions to review their assessments and determine if they can be easily completed using AI. This would require staff training in the potential and limitations of genAI.
According to the report, students are primarily using genAI to explain concepts, summarize articles, and generate research ideas. However, nearly 1 in 5 students admitted to including AI-generated text directly in their work.
While students find genAI helpful in saving time and improving the quality of their work, they are also concerned about being accused of academic misconduct and obtaining false or misleading results.
The report also noted differences in attitudes towards genAI among students, with women expressing more concerns and men showing greater enthusiasm. Students from wealthier backgrounds and those studying science, technology, engineering, and maths (Stem) courses also displayed greater enthusiasm for AI tools.
The digital divide appears to have widened, with half of students from privileged backgrounds using genAI to summarize articles, compared to 44% from less privileged backgrounds.
Students generally believe that their universities have effectively addressed concerns about academic integrity, with 80% stating that their institution’s policy is “clear” and 76% believing that their institution would detect the use of AI in assessments. However, only a third of students have received training in AI skills from their university.
Dr. Thomas Lancaster, a computer scientist at Imperial College London, stated that students not using genAI tools are now in a minority. He acknowledged concerns about the use of AI, but emphasized that students who reject AI may be at a competitive disadvantage in education and their future careers.
A spokesperson for Universities UK acknowledged the challenges posed by AI tools in exams and assessments. They stated that universities are actively educating students about the implications of cheating and have strict codes of conduct that include severe penalties for submissions of work that is not the student’s own.
Source: https://www.theguardian.com/education/2025/feb/26/uk-universities-warned-to-stress-test-assessments-as-92-of-students-use-ai
