Most AI acceptable use policies still look like basic HR housekeeping: a short list of approved tools, a caution about confidential data, and a signature line.
What they rarely address is the legal distinction that now carries practical weight in the EU — the difference between employees simply using AI tools and AI systems being used to make decisions about employees.
One sits under the Article 4 literacy duty (already enforceable). The other falls into Annex III high-risk territory (deferred until December 2027).
A practical template and supporting briefing have been published that treat this boundary properly:
→ Compliance briefing (PDF)
https://aigovernancedesk.com/wp-content/uploads/2026/09/employee-ai-use-policy-compliance-briefing-eu-ai-act-2026.pdf
→ Ready-to-adapt policy template (.docx)
https://aigovernancedesk.com/wp-content/uploads/2026/09/ai-acceptable-use-policy-template-for-employees-2026.docx
The template covers tiered tool approval, data classification rules tied to GDPR triggers, human-review requirements linked to legal effect, and a structure designed to serve as Article 4 evidence rather than a static onboarding PDF.
For teams responsible for AI governance, HR policy, or compliance in organisations operating in the EU, the materials are available for immediate use.