Key takeaways
- Check your employer's AI policy before anything else.
- Start with one assistant and six everyday tasks; don't collect tools.
- Good prompts include context, task, input, format and guardrails.
- Never paste confidential or personal data into tools your organization hasn't approved.
Step 1: Check the rules
Many organizations now have an AI use policy that lists approved tools and banned data. If yours does, follow it. If it doesn't, ask your manager or IT team which tools are allowed. Using an approved business version usually means your data is not used to train the provider's models.
Step 2: Pick one assistant
ChatGPT, Claude, Gemini and Microsoft Copilot are all capable general assistants. The best choice is usually the one that fits the software you already use: Copilot if you work in Microsoft 365, Gemini if you work in Google Workspace. Stick with one for a month so you learn its strengths instead of comparing tools.
Step 3: Start with six high-value tasks
- Email: shorten, clarify or adjust the tone of a draft.
- Meetings: turn notes or a transcript into decisions and action items.
- Research: get a quick briefing on a topic, then verify the key facts.
- Documents: turn rough notes into a structured first draft.
- Spreadsheets: get help with formulas, cleaning data or explaining a table.
- Learning: ask it to explain something at your level, then quiz you.
Step 4: Use a simple prompt formula
Think of the AI as a smart new colleague who knows nothing about your situation. Give it five things:
| Part | Example |
|---|---|
| Context | “I manage a five-person customer support team at a software company.” |
| Task | “Write a short update announcing our new weekend coverage rota.” |
| Input | “Here are the shift times and names: …” |
| Format | “Under 150 words, friendly, with the rota as a bullet list.” |
| Guardrails | “Don't promise extra pay. Ask me if anything is unclear.” |
If the first answer misses, give feedback rather than starting over: “Shorter, and put the start date first.” Our prompt library has 30 ready-made examples.
Step 5: Check before you use
AI assistants are fluent, not infallible. Before you send or publish anything, check names, numbers, dates, quotes and claims against a reliable source. Be especially careful with anything legal, medical, financial or safety-related, and remember that you are responsible for what you send.
Step 6: Protect sensitive information
Unless a tool is approved for it, don't paste in:
- Customer or patient personal data
- Passwords, keys or security details
- Unreleased financial results or confidential strategy
- Anything covered by a non-disclosure agreement
Step 7: Build the habit, then measure
Use AI for at least one real task every working day for two weeks. Keep a short log: the task, the time it used to take, the time it takes now and whether the quality held up. After two weeks you'll know where AI helps you most, and you'll have numbers to share with your manager.
Frequently asked questions
Is it cheating to use AI at work?
Not if your employer allows it and you are open about it where it matters. Using AI to work faster is like using a spreadsheet instead of a calculator. Passing off unchecked AI output as your own careful work is the problem.
Which AI tool should a beginner use?
Any major assistant works. Choose the one your organization approves or the one built into the software you already use, and focus on learning to prompt and verify well.