Skip to content
humanise.ai Forum

Is it okay to use AI to write work emails and reports?

Staff answerWritten by Bronwyn Ashby · AI at work and in the office1 replyupdated

Short answer

Often yes, but your employer's policy comes first, and so does what you paste in. Do not put personal, confidential or commercially sensitive information into a public AI tool. Use it for drafting and tidying, then check every fact, figure and name yourself, because you are the one who signs off the email or report.

Written by Bronwyn Ashby · Detection and policy editor ·
Citehumanise.ai Forum, "Is it okay to use AI to write work emails and reports?", https://humaniseai.ai/forum/read/is-it-okay-to-use-ai-to-write-work-emails-and-reports/, accessed 2026-11-20.
  1. Detection and policy editor ·

    Start with your employer's rules

    Workplaces differ more than universities do. Some have approved tools, some ban public chatbots outright, and some have said nothing yet. Before you use AI for anything at work:

    1. Search the staff handbook or intranet for "AI", "generative AI" or "acceptable use".
    2. Check your IT or information security policy, which often covers what you can upload to outside services.
    3. Look at any client contract or confidentiality agreement you work under. These can restrict where client material goes, whatever your employer's general AI policy says.
    4. If nothing is written down, ask your line manager, and ask for the answer in writing. "Is it fine to use [tool] to draft internal updates, if I leave out names and client details?" is a good, specific question.

    If your organisation provides an approved AI tool through a business account, use that, not a personal account. Settings and data handling for business accounts are often different from free consumer ones, and your IT team can tell you how.

    The part regulators care about: what you paste in

    Two regulators publish guidance that is directly relevant, one for each side of the world this forum serves.

    In Australia, the Office of the Australian Information Commissioner has guidance on privacy and the use of commercially available AI products. As a matter of best practice, it recommends that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools, because of the significant and complex privacy risks. It also says the Privacy Act applies to all uses of AI involving personal information, and that privacy obligations apply to personal information put into an AI system as well as the output it generates.

    In the UK, the Information Commissioner's Office has a hub of AI and data protection guidance, covering how to apply UK GDPR principles to AI systems and how to explain decisions made with AI. The ICO notes that, because the Data (Use and Access) Act came into law on 19 June 2025, this guidance is under review and may change, so check the page before relying on a detail.

    Neither regulator says "never use AI at work". Both are about personal information and how organisations handle it. As an individual employee, the practical reading is the same.

    A simple traffic light for what to paste

    Colour Examples Rule of thumb
    Green Your own rough notes with no names, a generic template, a public press release, a question about wording Usually fine in a permitted tool
    Amber Internal plans, draft figures, anything about a named colleague or client Only in an approved tool, and only if policy allows. Remove names and identifiers.
    Red Customer or patient details, staff records, health information, passwords, unpublished financials, contract terms under NDA Do not paste into a public tool

    Removing a name is not always enough. A job title, a date and a location can identify a person. If you would not be comfortable emailing the text to a stranger, do not paste it.

    What AI is good for at work

    • Turning bullet points into a first draft of a routine email.
    • Changing tone, for example firmer but polite, or shorter.
    • Restructuring a long report so the recommendation comes first.
    • Producing a clearer summary of your own meeting notes (without pasting names or sensitive items).
    • Suggesting subject lines, headings and an agenda.

    What you must still do yourself

    Check every fact. AI invents figures, dates and sources confidently. The same discipline applies as in the thread on checking that references from ChatGPT exist: if you did not see it in a real source, do not send it.

    Read it as your reader will. An email that is smooth but vague, or that reads like a press release, damages trust. The tips in how to make AI-assisted writing sound like me work just as well for a note to your team.

    Take responsibility. If the draft promises something your organisation cannot deliver, the promise is yours, not the tool's.

    Follow disclosure rules. Some employers and clients require you to say when AI helped with a deliverable. If yours does, say so in the form they ask for.

    If your organisation has no policy

    Ask for one, and meanwhile keep to the cautious end: no personal or confidential data, nothing in regulated areas such as legal advice, medical or financial decisions without sign-off, and a human read of everything before it goes out. Write down what you used it for, in case anyone asks. A few lines in your notes are enough.

    1
    ReplyLink

Plain text or simple Markdown. Links are nofollow. Your email is never shown.

  1. Harriet PellStaffForum editor, voice and editing ·

    If you are on placement or an internship, this applies to you too, and arguably more strictly. You may be under a confidentiality agreement you signed on day one. Ask your supervisor before using any AI tool on work material, and never assume your university's AI rules carry over to the workplace.

humanise.ai Forum. Questions and answers on using AI writing tools well at university and work, voice and tone, and AI detector worries.