AI Policy for Employees: A Starting Point for HR

Three Decisions to Make Before You Write
Get IT, security, legal, and a business leader in a room for an hour and agree on these:
1. Which tools are approved? List specific tools and what they’re approved for. Decide how employees can request a new one. A short approved list with a clear request path works better than a long banned list.
2. What data is off limits? Most companies restrict customer personal data, employee personal data, confidential financials, unreleased product details, credentials, and anything under a contract with confidentiality terms. Be specific about what can and can’t go into which tools.
3. Which decisions must stay human? Typical examples: hiring and firing, compensation, performance ratings, legal or medical advice, and anything published externally without review. AI can help prepare, but a person decides.
Once these are settled, the policy mostly writes itself.
An AI Acceptable Use Policy Template
Adapt this to your company. Replace bracketed text.
---
[Company] AI Use Policy
Purpose. We want everyone at [Company] to use AI to work better and grow faster, while protecting our customers, our colleagues, and our company. This policy explains how.
Scope. This policy applies to all employees and contractors using AI tools for [Company] work, on any device.
Approved tools. You may use the following tools for the purposes listed:
[Tool]: [approved uses, e.g. drafting, summarizing internal documents]
[Tool]: [approved uses]
[AI coaching tool]: personal growth, preparing for work conversations, goal-setting
To request a new tool, contact [team/channel]. Don’t use unapproved tools for [Company] work.
Data rules.
Never enter: [customer personal data, employee personal data, credentials, confidential financials, unreleased product information].
Internal, non-confidential information may be used in approved tools.
When in doubt, leave it out and ask [team/channel].
Human review.
You are responsible for any work you produce with AI, including accuracy and tone.
Review AI output before sharing it with customers or publishing externally.
AI may help you prepare, but the following decisions must be made by a person: [hiring, termination, compensation, performance ratings, legal matters].
Transparency. Tell people when content they rely on was substantially generated by AI, where it matters (for example, customer-facing material or formal documents).
Fairness. Don’t use AI in ways that could discriminate against candidates, employees, or customers. Flag any output that seems biased.
Questions and concerns. Contact [name/channel]. You will never be penalized for asking whether something is allowed.
Review. This policy is reviewed every [six] months.
---
AI Guidelines for Employees: The One-Page Version
Most people won’t read the full policy. Give them a short version alongside it:
Do:
Use approved AI tools to draft, summarize, brainstorm, and learn
Use AI coaching to prepare for conversations and work on your goals
Check every output before you share it
Ask if you’re not sure
Don’t:
Paste customer or employee personal data into AI tools
Use unapproved tools for company work
Let AI make decisions about people
Publish AI-generated content externally without review
How to Roll Out Your AI Policy
Lead with why. Open the announcement with what the policy helps people do, not what it forbids.
“We want everyone to get the most out of AI at work, safely. Here’s a short guide to which tools to use, what to keep out of them, and where to ask questions.”
Show examples. Give two or three examples per team: “Sales can use [tool] to draft follow-up emails from call notes. Don’t paste in customer contract terms.”
Brief managers. They’ll get the questions first. Give them a short FAQ.
Make asking easy. A dedicated channel for “can I use AI for this?” questions catches problems before they happen.
Offer learning. A policy works best alongside a little training. See how to learn AI as an HR professional for ideas you can adapt for other teams.
Keeping the Policy Current
AI tools change fast. Build in a review rhythm:
Review the approved tool list quarterly
Revisit the full policy every six months
Track the questions coming into your AI channel. Repeated questions point to unclear sections.
Update examples as new use cases emerge
For a broader view of where AI fits across the People function, see how to use AI in HR without losing the human part.
Related guides:
How Blomma Helps People Teams
An AI policy is more useful when the approved list includes tools that help people grow, not just tools that help them work faster. Blomma gives every employee a private AI career coach -- a thought partner for feedback, hard conversations, goals, and career growth that pairs AI with emotional intelligence.
It’s a natural fit for the “approved tools” section: purpose-built for coaching, private between each person and their coach, and available at any hour. Instead of employees turning to random tools for advice about a tough situation at work, they have an approved coach built for exactly that. For the People team, it’s a way to show your AI policy is about support, not only rules.
