AI acceptable use policy training # Everyone signed the AI policy. *Can they apply it?* Short modules on the moments a policy is actually tested: the paste into the wrong account, the confident wrong answer, the document that gives an assistant orders. Then a versioned acknowledgement of your own policy, and a dated record of who is current. [Book a demo](https://trainspace.dev/contact) [See the modules](https://trainspace.dev/ai-acceptable-use-policy-training#modules) Priced per seat, per year. The short answer ## How do you train employees on an AI acceptable use policy? Teach the few decisions the policy governs, each with a check: which tool and account to use, what data may go in, what to verify before acting, when to say AI helped and who to tell when it goes wrong. Then have everyone acknowledge the current policy version. Train Space does both and keeps the record of who did what, and when. - **Modules:** 11, from three programs - **Your policy:** Versioned, acknowledged by everyone - **Pass mark:** 80%, scored on the server - **Record:** Who is current as of any date - **Facts checked:** 2026-11-04 What people learn ## The policy, *one decision at a time.* Each module ends in a check. The outcome column is what the person can do afterwards. | Module | What the person can do after it | Program | | --- | --- | --- | | What may be pasted into which tool? | Tier what you're about to paste in under ten seconds, send it only where that tier may go, and cut it down so a lower tier applies. | AI essentials | | Why does it matter which account you paste work into? | Check the account before you paste work into an AI tool, know what a work account changes, and tell IT when it falls short. | AI essentials | | Using approved software and AI tools | Check a tool is approved, use it on your work account with only the data its label allows, and ask IT when it falls short. | Security Awareness Program | | How can a document someone sends you take over your assistant? | Spot a file steering your assistant, avoid set-ups that mix private data, outside content and a way to send, and approve every send, share and payment yourself. | AI essentials | | AI assistants and hidden instructions | Spot when an AI assistant takes orders from what it reads, approve every send, share and payment yourself, and report the planted note. | Security Awareness Program | | How do you tell an answer is wrong before acting on it? | Name an AI answer's most likely mistake, run the matching check in under a minute, and stop an unsupported claim before anyone else sees it. | AI essentials | | What does human oversight actually require? | Test whether the person approving an AI's output can change it, show them the evidence, and prove the check happens. | AI essentials | | When and how should you tell people AI helped? | Add one plain line to AI-assisted work someone will rely on: what AI did, what you checked and what you didn't, within your organisation's rules. | AI essentials | | When does connecting one more app make your assistant dangerous? | Count the three legs of risk in each assistant task, remove one before it goes live, and read approval prompts properly. | AI on real work | | What should you check before installing a skill or plugin someone else wrote? | Check a skill or plugin's maker, reach, instructions and approval before you switch it on, and again after it updates. | AI on real work | | Which AI uses at work carry legal obligations? | Spot AI uses that decide people's jobs, money or services, and take them to legal with the right questions before launch. | AI on real work | How it runs ## From a PDF nobody reads *to a record you can show.* AI tools add features every month, and a policy written in spring can miss what shipped in autumn. The modules are rechecked against vendor sources on a schedule, so the habits people learn match the tools on their screen. 01 LOAD ### Your policy, versioned. Add the policy as a PDF or a link and mark it required. Every new version starts a new acknowledgement. ✓ Your words, not ours 02 TEACH ### The decisions it governs. Short modules on tools, accounts, data, checking and disclosure, using your approved tools and channels. ✦ Checked, not just viewed 03 RECORD ### Who is current, any day. Completion, scores and acknowledgements by date, ready to export when someone asks. ↻ Kept current as the tools change Who it is for ## Everyone with an AI tool. *Which is now everyone.* 01 / IT AND SECURITY ### Approved tools, used properly. People learn to check a tool is approved, use the work account and report a slip the moment it happens. 02 / LEGAL AND GRC ### A dated acknowledgement. Each policy version has its own acknowledgement and due date. Department lenses rewrite scenarios for legal and GRC teams. 03 / MANAGERS ### Who is current, at a glance. See who has done which modules and acknowledged which version, as of any date you pick. A few good questions ## Rolling out an AI policy? *Make it stick.* Still wondering about something? Ask us and we’ll answer it here. [Ask your question](https://trainspace.dev/contact) ### ### What is AI acceptable use policy training? Training that turns a company's AI acceptable use policy into habits people use every day: which AI tools are approved, what data may go into which tool, using a work account rather than a personal one, checking answers before acting on them, saying when AI helped, and reporting when something slips. It ends with each person acknowledging the current version of the policy. ### ### Isn't sending the policy round enough? A signed policy shows people saw it, not that they can apply it. The usual failures are practical: pasting a client file into a personal account, trusting a confident wrong answer, or an assistant following instructions hidden in a document. Each of those is a short module here, with a check scored on the server. ### ### Can we use our own AI policy? Yes. Keep your policy in Train Space as a PDF or a link, mark it required, and every seated person acknowledges its current version, with their name, email and time recorded. Publish a new version and everyone owes an acknowledgement of it, with its own due date. Lessons name your approved tools and reporting channel through your company details. ### ### What should an AI acceptable use policy cover? At least: which tools are approved and on which accounts; what data may go into each tool; what must be checked by a person before it is used or sent; when to say AI was used; what agents and plugins may connect to; and who to tell when something goes wrong. Ask your counsel about the rules that apply where you operate. ### ### Which modules does it use? 11 modules from AI essentials, AI on real work and the Security Awareness Program, listed on this page. Each one names the sources it was checked against and the date, and a lesson past its recheck date can't ship. ### ### How do we show who has done it? Managers see who is current as of any date. The security program's audit export lists, for each person, the cycle, module scores, the content version and when they acknowledged each policy version, as CSV or JSON with a SHA-256 hash of the CSV. Keep exploring [### AI training for employees Three programs, from first prompt to agents, with scored checks. Read more](https://trainspace.dev/ai-training-for-employees) [### Security awareness training The yearly program behind the policy acknowledgements and audit export. Read more](https://trainspace.dev/security-awareness-training) [### Secure AI coding training For developers using Claude Code, Copilot and Cursor. Read more](https://trainspace.dev/secure-ai-coding-training) Put the learning to work ## Turn the policy into habits. *And keep the proof.* Book a demo and we'll show you the modules, the acknowledgement and the record. [Book a demo](https://trainspace.dev/contact) Priced per seat, per year.