The English higher education regulator, the Office for Students, has warned universities that they may need to replace traditional at-home essays and open-book exams with supervised assessments that make it harder for students to rely on artificial intelligence when producing work.
The agency's report states that AI detection tools are too unreliable to consistently determine whether students have used the technology. Commercial AI detection tools used to assess writing are far from perfect and can produce false positive results, the document warns. AI systems are also becoming increasingly sophisticated, making reliable detection even more difficult.
As a result, universities are increasingly turning to in-person written and oral exams, which significantly make it harder for students to outsource work to AI, the report states. The authors also pointed to process-based assessments, which evaluate students at multiple checkpoints, such as literature reviews, rather than solely on the basis of final essays.
The report illustrates how widespread AI use has become in universities and why lecturers are increasingly facing the need to fundamentally rethink how student success is assessed, sometimes with unintended consequences. It highlighted concerns that some secure assessments, such as unannounced exams, bear little resemblance to the activities students will undertake in professional environments.
Particular concern was raised that structured formats and rewarding long-term memory could disproportionately disadvantage certain groups of students, such as those with disabilities. Research has also found that people find it difficult to distinguish between AI-generated images and real objects.
The report also mentions concerns that a gap could emerge between students whose parents can afford expensive paid AI tools and those forced to rely on cheaper or free models. There are fears that using AI for schoolwork could become a gateway to over-reliance on chatbots, particularly for emotional support.













