OfS warns AI makes grading coursework 'close to impossible'

England's higher-education regulator has told universities that artificial intelligence has made grading unsupervised coursework 'close to impossible', and warned that institutions failing to design reliable assessments could face sanctions. The joint report from the Office for Students and Advance HE, titled *Making AI Work for Students* and published on 11 November, draws on a four-month investigation covering roughly 300 students and staff across more than 120 institutions.

The sharpest finding is in the students' own words. Interviewees described peers who 'completed assessments using AI, sometimes in just a few minutes, and received a similar grade to those who had completed the assessment themselves'. Some student participants said the assessments run by their institution are 'not fit for purpose or do not accurately measure students' knowledge and skills'. The report is blunt about the technical fallback too: 'AI-detection tools do not consistently and reliably identify AI use'.

Josh Fleming, the OfS's director of strategy and delivery, framed the issue in generational terms. AI, he said, presents a 'generational risk to the system as it stands, threatening to undermine students' learning and institutions' ability to assess them', and he called for 'far greater use of AI-free assessment'.

The regulator's practical ask is a shift to formats AI cannot sit on a student's behalf: in-person examinations, presentations, vivas, and performance-based or practical tasks. Its leverage is the conditions of registration: institutions that fail to produce valid and reliable measures of student knowledge could be in breach, with sanctions in prospect. Two researchers on our radar shared the story on the day it landed.

The report does not name any individual institution, and it does not set a deadline by which universities must redesign their assessment regimes.