A problem for people who use AI
Connecting AI learning to the practice of each role and major
Administrative and management staff at universities and colleges, university students, and government staff
Training and consultation carried out; effect unmeasured
The meeting was October 2026. The student survey was December 2025 and the US staff survey is 2025 background material.
Editorial check: 2026-10-23
What people are paying attention to
At the October Jisc meeting, administrative staff at universities and colleges discussed AI use and support. The background survey of students also shows interest in AI skills that connect to work.
What is hard
Workplaces need different ways of using it by department and role, and people also ask for places to consult. The student survey shows how support is perceived, but the form needed is not settled.
Workarounds being tried
The meeting reported learning tasks tried in the workplace and consultation support. The student survey's authors also propose introductory education and building it into major-specific classes.
What the sources confirm
Institutions trying training and consultation have reported, but no results measuring recipients' skills or effects on work are shown. The student survey also did not test actual skills.
What is still unclear
Whether people who received support became able to use AI on the tasks of their own role or major and judge the output is unconfirmed. Making concrete the support users want also remains an open question.
Workarounds and results in more detail
Combine workplace dialogue with small practice
Examples of implementation reported at the meeting
Shared short learning tasks, practice making meeting materials, and training, rules and consultation support. No measurement of effect is shown.
Scope and limits: The organizer summarized the experiences of institutions that took part in the Jisc meeting. The number of participants and the institutions' locations are unconfirmed.
Provide role-specific training and a common consultation desk
Request from users
No result of introducing or evaluating support in response to this request can be confirmed in the source.
Scope and limits: Staff of the North Carolina Department of Transportation answered after the 2025 trial.
Prepare introductory education and major-specific skills education
Proposal by the authors; effect untested
They recommend learning support and assignment-specific usage guidance. Changes in skills from these measures were not measured.
Scope and limits: A proposal by HEPI authors responding to the December 2025 UK undergraduate survey.
Read with care
- The meeting is the organizer's summary, and the representativeness of participants and the effect of training are unconfirmed. It cannot be generalized to workplace trends overall.
- The student survey is self-reported, and current university support may not be the same. Kortext is a sponsor.
- Students, administrative staff and transportation department staff are different groups. Shortfalls or effects of support cannot be shown with a common rate.
What to check next
Records for the same role and major of handling real tasks before and after support, what they consulted about, and the uses they kept and dropped.
Staff bring everyday materials together and share ways of using and judging
At a Jisc meeting on administrative staff in education, a situation came up in which people in the same department did not know how others used AI. Besides basic training, opportunities to talk about use in teams, short learning activities, and practical tasks such as making meeting outlines and agendas were introduced. Judgments on information handling and approval are also needed, and the issue is a route to consult when staff are unsure.
In HR, finance, student support and similar areas, the information handled and needed outcomes differ, and it was discussed that even after a course people need to relearn uses and ways of checking. What can be confirmed here is the introduction of participating institutions' practices and opinions at the meeting. The effect and continuation rate of training are not measured. Editorially, this is treated as a case that checks together a place to try with one's own materials and whom to consult when judgment is unsure.
Questions to think with
- Among the materials a staff member actually makes, which information can be handed to AI and which content do they judge themselves?
- Is there someone to consult about where a trial stopped, and a chance to redo it with other material?
Editing that checked the primary sources. The questions to check in your own work are proposals for verification.