AI

Staff guide to using AI responsibly

AI can be used in day-to-day work as a tool to help with emails, meeting summaries, first drafts, summarising reports, identifying themes, and improving tone and clarity. Here are some do's and don'ts to help guide you.

Do's

  • Do start with approved tools
    Use University-supported AI tools, as they’re assessed for safety and data protection.
  • Do check your data classification first
    Stick to public or internal data unless you have explicit approval.
  • Do be clear in your prompts
    The better your instruction, the better the output; add context, tone, and purpose.
  • Do treat outputs as a first draft
    Always review, edit, and sense-check before using or sharing.
  • Do stay accountable
    You are responsible for anything produced using AI.
  • Do use AI to support, not replace, your judgement
    Keep a human in the loop for decisions and final outputs.
  • Do be transparent when appropriate
    Let others know if AI has supported your work, especially in formal outputs.
  • Do think about bias and inclusion
    Check outputs for tone, assumptions, and fairness.
  • Do use AI to save time on routine tasks
    Drafting, summarising, structuring, and idea generation are strong use cases.

Don'ts

  • Don’t input sensitive or personal data
    Avoid confidential, personal, or restricted information unless approved.
  • Don’t assume outputs are correct
    AI can be wrong, outdated, or misleading. Always verify.
  • Don’t rely on AI for final decisions
    Especially where there are risks, impacts, or policy implications.
  • Don’t use unapproved tools for University work
    Even if convenient, they may pose data or compliance risks.
  • Don’t copy and paste without reviewing
    Check for accuracy, tone, and appropriateness.
  • Don’t overuse AI where it adds little value
    Use it where it genuinely improves efficiency or quality.
  • Don’t ignore intellectual property considerations
    Be mindful of originality and attribution.
  • Don’t present AI-generated content as fully your own without review
    You remain responsible for quality and integrity.
  • Don’t assume AI understands context
    It doesn’t know your organisation so guide it clearly.
  • Don’t feel you have to use AI for everything
    It’s only a tool, not a requirement.

Related