Free staff training
Safe AI at Work: free AI training for your whole team
25 short, plain-English modules that help non-technical staff use AI tools productively while they mitigate the risk to your business. Free from DAS CX. Share it with everyone.
Modules25
Time per module3–5 minutes
FormatQ&A with quick checks
CostFree
Why it matters
Your staff are already using AI. Do you know how?
We recently sat down with the leadership team of a fast-growing business. As far as they knew, AI wasn’t really being used yet. Then we looked at how people actually worked. Staff across the business were using ChatGPT, Claude and other tools every day, mostly free versions on personal accounts. Some were uploading confidential documents.
Nobody was being careless on purpose. They were trying to do their jobs faster. But free consumer tools may keep what you put in, and once information leaves your business you can’t take it back. In 2023, Samsung engineers pasted confidential source code into ChatGPT, and Samsung restricted AI tools across the company.
Blocking AI isn’t the answer. People just use it on their phones, with even less visibility. The answer is to keep the productivity and mitigate the risk.
Keep the productivity. Mitigate the risk.
- Build awareness. Help people understand what’s safe to share and why it matters.
- Provide safer tools. Give staff business-grade AI with proper data protection.
- Teach safer methods. Remove names, client details and financials before using AI.
- Set clear guidelines. A simple, plain-English acceptable-use policy.
- Keep listening. Find out where AI is helping, then build on it.
This free training covers the first and third steps for every employee.
How to use it
Share it with all staff
Ask everyone to complete one or two modules a day and finish within two weeks. Managers can use the quick checks as a short quiz.
Part 1 · Modules 1–4
AI Basics
What these tools are, why your business wants you to use them and where your information goes.
Module 1What is generative AI?
What do tools like ChatGPT, Claude, Copilot and Gemini actually do?
They predict and generate text, images or summaries based on what you type in (your prompt). They’re very good at drafting, summarising and explaining, but they don’t “know” things the way a person does.
Are they always right?
No. They can sound confident and still be wrong. Treat AI output as a first draft, never a final answer.
Quick checkTrue or false: if AI sounds confident, it’s accurate.
Show answer
False
Module 2Why your business wants you to use AI
Is using AI at work allowed?
In most businesses, yes, as long as you follow the company’s guidelines and use approved tools. AI can save hours on routine work.
What’s the catch?
What you type into an AI tool may leave the business. Using AI well means knowing what’s safe to share.
Quick checkWhat’s the main goal?
Show answer
Keep the productivity, mitigate the risk.
Module 3Where does my information go?
When I paste something into an AI tool, where does it go?
To the provider’s servers. Depending on the tool and settings, it may be stored, reviewed by people, or used to train future models.
Can I delete it later?
Sometimes you can delete your chat history, but you can’t guarantee it’s gone from every system. Assume anything you share could be kept.
Quick checkOnce data leaves the business, can you always get it back?
Show answer
No.
Module 4Free tools vs business tools
What’s the difference between free ChatGPT and a business version?
Business and enterprise versions usually come with contracts that say your data won’t be used for training, plus admin controls and better security. Free and personal accounts often don’t.
Which should I use for work?
Always the tools your company has approved. If none are approved yet, ask before using AI for work tasks.
Quick checkIs a personal free account okay for client work?
Show answer
No.
Part 2 · Modules 5–9
What Never Goes Into AI
The information that should never be pasted or uploaded into an AI tool.
Module 5Personal information
What counts as personal information?
Anything that identifies a person: names, addresses, phone numbers, emails, dates of birth, photos, and ID numbers like tax file, licence or passport numbers.
Can I paste a customer email into AI to help me reply?
Only in an approved business tool, and remove names and contact details first. Better still, describe the situation instead of pasting it.
Quick checkIs an email address personal information?
Show answer
Yes.
Module 6Sensitive information
Is some information even more sensitive?
Yes. Health, financial, legal, HR and performance details, and anything about children, need extra care. Never put these into AI unless your company has specifically approved it.
Quick checkCan you paste a staff performance review into a free AI tool to “tidy it up”?
Show answer
No.
Module 7Client and customer data
What about client information?
Client names, contracts, pricing, account details and project information belong to the client relationship. Sharing them with an outside tool can breach contracts and trust.
Quick checkWho could be harmed if client data leaks?
Show answer
The client, the business and you.
Module 8Company confidential information
What’s company confidential?
Financials, strategy, board papers, pricing models, unreleased products, internal reports, source code and anything marked confidential.
But I only want a summary.
The tool still receives the full document. Summarising doesn’t make sharing safe.
Quick checkDoes asking for “just a summary” protect the document?
Show answer
No.
Module 9Passwords and access details
Can I ask AI to help with a login or system problem?
Yes, but never include passwords, API keys, access codes, account numbers or screenshots that show them.
Quick checkShould a screenshot with a visible password go into AI?
Show answer
Never.
Part 3 · Modules 10–14
Spotting Risk Before You Upload
Quick habits that catch risky information before it leaves the business.
Module 10The 5-second check
How do I quickly decide if something is safe to share?
Before you paste or upload, ask:
- Does it name a real person?
- Does it include client or customer details?
- Is it confidential or internal only?
- Would I be comfortable seeing this on a public website?
- Am I using an approved tool?
If any answer worries you, stop and remove the detail or ask.
Quick checkWhat’s the “public website” test?
Show answer
If you wouldn’t want it public, don’t share it.
Module 11Hidden data in documents and files
Can a file contain information I can’t see?
Yes. Spreadsheets can have hidden tabs and columns. Documents can have tracked changes, comments and metadata. PDFs and images can include names and details you’ve forgotten about.
Quick checkName one place hidden data hides.
Show answer
Hidden tabs, comments, tracked changes, metadata.
Module 12Screenshots and images
Are screenshots safer than copying text?
Not really. Screenshots often capture names, emails, open tabs and notifications. Crop tightly and check every corner before uploading.
Quick checkWhat should you do before uploading a screenshot?
Show answer
Crop and check it.
Module 13How to anonymise information
How do I use AI without sharing private details?
Replace real details with placeholders: “Client A” instead of a company name, “[Name]” instead of a person, rounded numbers instead of exact figures. Keep the structure, remove the identity.
Quick checkRewrite “Email John Smith at Acme about the $48,250 invoice.”
Show answer
“Email [Name] at Client A about the invoice.”
Module 14Describe, don’t upload
What’s the safest way to get help?
Describe the problem in general terms. “How do I politely decline a refund request?” works as well as pasting the actual customer email, with none of the risk.
Quick checkWhich is safer: uploading the contract or asking how to structure a contract clause?
Show answer
Asking.
Part 4 · Modules 15–20
Using Common Tools Safely
How to use ChatGPT, Copilot, Claude, Gemini and built-in AI features safely at work.
Module 15ChatGPT
How do I use ChatGPT safely?
Use your company’s approved account if there is one. In personal accounts, check Settings → Data Controls and turn off model training if allowed. Never share personal, client or confidential information.
Module 16Microsoft Copilot
Is Copilot different?
Copilot in a company Microsoft 365 account works within the business’s own security, and can see files you have access to. That’s powerful, so keep file permissions tidy and don’t use the free consumer version for work.
Module 17Claude, Gemini and others
What about other AI tools?
The same rules apply. Know whether you’re on a business or personal account, check data settings, and only use tools your company approves.
Module 18AI inside other apps
Is AI showing up in apps I already use?
Yes. Email, meeting, design and note-taking apps now have AI built in. These features may send your content to outside providers. Check with IT before switching them on for work.
Quick checkIs a built-in AI feature automatically approved?
Show answer
No, check first.
Module 19AI meeting recorders and note-takers
Can I use an AI note-taker in meetings?
Only if your company approves it and everyone in the meeting knows and agrees. Recordings can capture confidential and personal information.
Quick checkWhat must you do before recording a meeting with AI?
Show answer
Get approval and tell everyone.
Module 20Browser extensions and unknown apps
Are AI browser extensions safe?
Be careful. Some extensions can read everything on your screen, including emails and internal systems. Only install what IT has approved.
Part 5 · Modules 21–23
Checking What AI Gives You
Why AI output is a first draft, never a final answer.
Module 21Hallucinations
What’s a hallucination?
When AI makes something up, like a fake statistic, quote, law or reference, and presents it as fact. Always check facts, numbers and sources before you use them.
Quick checkAI gives you a statistic. What do you do?
Show answer
Verify it from a trusted source.
Module 22Bias and tone
Can AI be biased?
Yes. It learns from internet data and can reflect stereotypes or an off-brand tone. Read everything as if your name is on it, because it is.
Module 23Copyright and ownership
Can I use anything AI creates?
Be cautious with images, logos and long passages that could copy someone else’s work. Don’t present AI output as original research, and follow your company’s policy.
Part 6 · Modules 24–25
Staying Safe Day to Day
AI-powered scams, and exactly what to do if something goes wrong.
Module 24Scams and AI-powered phishing
How are scammers using AI?
To write convincing emails, fake voices and even fake videos. Be suspicious of urgent requests for money, passwords or data, even if they sound like your manager. Verify through a separate channel.
Quick checkYour “CEO” emails asking for an urgent transfer. What do you do?
Show answer
Call them directly to confirm.
Module 25What to do if something goes wrong
I think I shared something I shouldn’t have. What now?
Don’t panic and don’t hide it. Delete the chat if you can, note what was shared and when, and tell your manager or IT straight away. Speed matters more than blame.
Quick checkWhat’s the most important thing to do after a mistake?
Show answer
Report it quickly.
Keep this handy
Your Safe AI cheat sheet
Do
- Use company-approved AI tools
- Remove names and client details
- Describe problems instead of uploading files
- Check facts and numbers
- Ask if you’re unsure
Don’t
- Share personal, client or confidential information
- Use personal free accounts for work data
- Paste passwords or access details
- Trust AI output without checking
- Hide a mistake
For businesses
Want to go further?
The free training builds awareness. These help you turn safe AI use into a real advantage.
1
On-site Safe AI workshops
Live training tailored to your approved tools, your roles and the information your business handles.
Training and change →2
AI acceptable-use policy
A one-page, plain-English policy that says what’s fine, what’s never okay and who to ask.
Ask us about it →3
Build Your AI Roadmap
Find out how AI is really being used across your business, where it could help most and how to bring your people with you.
The ALIGN framework →Safe AI at Work questions
Is the Safe AI at Work training really free?
Yes. All 25 modules are free to read and share with your whole team. There’s no sign-up. If you want more, we offer on-site Safe AI workshops, an AI acceptable-use policy and a full AI roadmap.
Who is it for?
Non-technical employees in any business: administration, operations, sales, finance, HR and customer service. Anyone who uses, or might use, AI tools like ChatGPT, Copilot, Claude or Gemini at work.
How long does it take?
Each module takes about 3 to 5 minutes and ends with a quick check. Most people complete one or two a day and finish within two weeks.
Can we use it in our staff induction?
Yes. Share the link with new starters, or print it and save it as a PDF. Managers can use the quick checks as a short quiz.
Does it replace an AI policy?
No. It teaches safe habits, but your business still needs a clear AI acceptable-use policy that names your approved tools. We can help you write a one-page policy in plain English.
Can you tailor the training to our tools?
Yes. Our on-site Safe AI workshops are tailored to your approved tools, your roles and the kind of information your business handles.
Make AI safe across your business
Book a free 30-minute AI Quick Scan. We’ll talk through how your people use AI today and the quickest ways to reduce the risk.