Safe AI at work,
in 5 minutes
Your team already uses AI. This short training teaches every employee how to keep company data out of it: what never gets pasted, which account to use, and what to do when a guardrail steps in.
Six short modules, one habit each
Start with the flagship. New deep-dive modules land weekly, each one under three minutes.
Safe AI at Work
The whole training in one sitting: three habits that keep company data out of AI tools.
Watch nowWhat Never Leaves the Building
Real leak stories for each data class, and the grey areas nobody warns you about.
Watch nowShadow AI and Approved Tools
Why the free new AI tool is not free, and how to champion a tool the right way.
Watch nowWhen the Guardrail Catches You
Redacted, blocked, or warned: what each means and why nobody is reading your chats.
Watch nowTrust, Verify, and AI-Made Deception
Hallucinations, deepfakes, and AI phishing: verify what AI gives you and what it sends you.
Watch nowYour AI Has Agents Now
Agents act on your laptop. What MCP tools are, and why agent permissions deserve the email test.
Watch nowCertificate of completion
Complete all 6 modules (watch + pass the quick check) to unlock your certificate. 0 done so far.
0/6
Three habits keep your data yours
No 40-slide course. These three habits catch almost every AI data leak before it happens.
The email test
Would you email it to a stranger outside the company? If not, it does not go into an AI tool either. That one question catches almost everything.
Work account for work
Your personal ChatGPT login puts company data where the company cannot see, protect, or delete it. Same tool, same answers: use the approved account.
Verify before you ship
AI answers sound confident even when they are wrong. Check numbers, names, dates, and code before they leave your hands. You own what you ship.
The never-paste list
- Passwords, API keys, tokens, anything that unlocks something
- Customer details: names, emails, order history, support threads
- Colleague information: salaries, reviews, anything HR
- Unreleased numbers: revenue, forecasts, deal terms
- Proprietary source code, unless the tool is approved for it
- Anything under NDA or marked confidential
Found a great new AI tool? Do not sneak it in on a personal account. Ask IT or security to approve it: at most companies that takes days, and then you can use it without thinking twice.
AI literacy is now a legal requirement
EU AI Act Article 4 has applied since February 2025, and national supervision begins on August 2, 2026. If your staff use AI systems at work, your organisation is a deployer.
- The European Commission is explicit: relying on an AI tool's instructions for use is not sufficient training.
- Measures must fit your organisation: your tools, your data, your roles. Generic e-learning alone does not meet the bar.
- You should be able to show who was trained, on what, and when: a completion record that stands up to an auditor.
The AI Safety Rollout Kit
Everything you need to roll out this training to your whole company this week, free.
- A copy-paste email to send your staff, with the video linked
- A 10-question quiz pack to confirm completion
- The one-page cheat sheet as a printable PDF
- An EU AI Act Article 4 evidence checklist
- A sample AI literacy evidence report, the kind auditors ask for
Training is step one.
Evidence is the platform.
Aona customers get this training assigned automatically, tracked per employee, with real-time redaction when someone pastes what they should not. One platform from awareness to audit evidence.