Data, privacy, copyright & accessibility (UK HE)

Who's it for?

Audience: staff and students who are about to paste something into an AI tool.
Purpose: tell you what’s OK to upload, what needs checking/anonymising, and what you should never put into a public AI.

Audience: Staff and students who are about to paste something into an AI tool.

Purpose: To tell you what’s OK to upload, what needs checking/anonymising, and what you should never put into a public AI.

Traffic Light Data Guide

🟢 GREEN – generally OK in public/chat AI

Your own writing (to improve clarity)
Assignment briefs (without internal staff notes)
Publicly available texts you’re analysing
Generic interview/STAR/Gibbs prompts
Practice questions

🟡 AMBER – check first / anonymise

Student work you are marking
Peer feedback / discussion board posts
Internal teaching materials
Case studies with some context
Drafts of research ideas
Placement scenarios (fictionalised)

👉 Prefer an institutionally approved / private AI for these, or strip names and identifiers first.

🔴 RED – don’t upload to public AI

Patient / service-user / school pupil data
Real placement / clinical notes
Personal data about named individuals
Commercially sensitive partner information
Unpublished research data with consent limits

👉 Use your university’s secure routes instead, or don’t use AI for that part at all.

Why this matters

Public/consumer AI tools may store or reuse what you type.
You may be under GDPR, ethics, PSRB or placement agreements that forbid sharing client/patient data.
Students often don’t realise their reflective pieces contain identifiable details.

So: if in doubt, don’t paste it.

Staff-Specific Guidance

When asking students to submit a process pack (prompts + AI output + reflection), tell them to anonymise any real-person/context data first.
If you want AI to help with feedback, paste your marking notes, not the full student script.
If your School has an enterprise / institutionally approved AI, link to it here and recommend that first.

Students on placement / Practice / Clinical Courses

Create a teaching version of your case before using AI (change names, locations, dates, identifiers).
Never share real patient/client/school records with a public AI.
If the task is assessed, ask your tutor which version to submit.
If AI helped you structure a reflection, say so (see How to acknowledge GenAI)

Copyright & Academic Integrity

Students:

AI may generate text/images that look original — you still need to cite real sources.
If AI “suggests” a source, check if it actually exists in the library/databases.
Saying “AI wrote it” does not make plagiarism OK.
Do not upload your taught material to public GenAI tools without staff permission

Staff:

Model best practice, if you generate teaching materials with AI, check for copied/attribution-required material.

Accessibility & Inclusion (The positive bits)

AI can be very helpful for:

simplifying assignment briefs (“explain this for a first-year student”)
rephrasing feedback into clearer English
giving alternative explanations for the same concept
drafting alt-text or stem sentences for students who need more structure

But, the official/assessed version of the brief/feedback is still the one from your module.

Local / Institutional Policy Comes First

This page is general UK HE advice. If your university / your School issues stricter rules, follow those.

See Core guidance → Local / institutional policy first for advice.

Page updated

Report abuse