OfS AI guidance connects feedback quality with clear student expectations

Updated Nov 12, 2026

Students need to understand both how they may use AI and how their tutors use it in feedback. On 11 November 2026, the Office for Students and Advance HE published Making AI work for students, alongside OfS examples of institutional practice under its existing quality conditions. For quality and Student Experience teams, the useful next step is to connect assessment guidance with the experiences students describe.

What students reported

The report draws on engagement with students and staff between June and September 2026, primarily concerning English higher education. Student participants described uncertainty about permitted AI use, including policies that varied between departments or staff. Some said they would avoid declaring permitted use because they feared it would cast suspicion on their work.

Feedback itself was another concern. Students criticised AI-assisted feedback when they considered it poor quality or felt staff had been unclear about its use. Staff described a possible role for AI as a starting point that could free time for direct work with students. The report brings those perspectives together around a practical question: what should expert feedback help the student understand and do next?

Students also asked for more opportunities to develop AI literacy, particularly in their subjects. Course teams can use that request to discuss where students find suitable learning support.

What the OfS has clarified

The new institutional practice page explains examples likely to comply with existing conditions B1, B2 and B4. These concern course quality and relevant skills, resources and support, and effective assessment respectively. Its direct regulatory scope is England.

For AI-assisted feedback, the OfS describes oversight that keeps feedback accurate, aligned with assessment criteria, tailored to the student's work and specific about improvement. It also describes clear, consistent guidance on permitted AI use and support that helps students understand the rules. The page gives examples of practice under existing conditions; the separate discussion of provider submissions is explicitly part of the proposed future quality system.

The accompanying announcement also stresses assessment validity. It calls for secure assessment where needed and for students to demonstrate the knowledge and skills being tested. That makes clear instructions one part of a wider assessment-design task.

A useful course-review task

Our recommendation is to choose an upcoming assessment and ask students to explain its AI instructions in their own words. Can they identify permitted activities, the declaration expected and where to seek an answer? Give the teaching team the specific wording that caused confusion and agree who will revise it. The assessment and feedback pathway connects this task with reviewing how students use feedback.

For feedback already returned, ask students to identify the criterion being discussed and the next action they would take. Keep a difficulty interpreting the advice distinct from a difficulty finding it or a question about AI use. Those are different decisions for teaching, digital-learning and quality teams.

Open comments can help locate these issues across modules. Preserve the assessment context, group related concerns and return to the source accounts before agreeing a response. Use the digital learning pathway to connect that review with the wider learning journey, then report the revised guidance and the question to be checked next.

FAQ

Q: Has the OfS introduced a new AI-specific condition?

A: The practice page explains examples under existing conditions B1, B2 and B4. It also identifies how assessment arrangements would be considered under the proposed future quality system.

Q: What should a course team review first?

A: Choose an assessment students will encounter soon. Ask them to explain its AI instructions, identify the help route and describe what they must declare. Assign an owner to resolve unclear wording.

Q: How can student comments inform a feedback review?

A: Separate comments about the advice itself, access to it and expectations about AI. Use those distinctions to choose a concrete response, then ask students to try the revised guidance or feedback task.

References

Office for Students and Advance HE: Making AI work for students: reflections on institutional practice, OfS 2026.56. Published 11 November 2026.

Office for Students: Effective institutional practice in AI. Published 11 November 2026.

Office for Students: Universities should help students develop AI skills while protecting confidence in degrees. Published 11 November 2026.