Free company AI policy template

Copy, adapt and share · Updated 8 October 2026

Give your team clear rules for using AI at work. Copy this one-page policy into your document, replace the brackets and agree it with the people responsible for AI in your company.

Download editable text Download one-page PDF

[Company name] AI acceptable use policy

Policy owner: [name or role and contact]
Approved by: [name or role] | Effective date: [date] | Next review: [date]
Approved tools and uses: [link to your AI tool register]

This policy applies to everyone using AI for [Company name] work. We use AI to support useful work while protecting company information and keeping people responsible for the results.

1. Approved tools and uses

Use only the tools, account plans and tasks in our approved AI tool register. Ask the policy owner before adopting a new tool, adding an integration or starting a paid subscription. Approval for one task or type of information does not cover every use.

2. Protect information

Never enter passwords, API keys or other access secrets. Use personal, client or confidential information only where the tool and task are explicitly approved for that type of information. Follow our existing privacy, security and client requirements. If unsure, use public or fictional material and ask before uploading.

3. Accounts and automated actions

Use company-approved work accounts and settings for company work. Do not use personal AI accounts. Get approval before giving an AI agent or integration access to company systems, and keep its access and actions within the agreed scope.

4. Check the result

The person using AI is responsible for the final work. Check important facts, sources, calculations, code and rights before relying on or sharing output. Consequential decisions about people need a separately approved process and accountable human review.

5. Be open about AI use

Disclose material AI assistance when it matters to the recipient, is required by a client, or falls under [company disclosure rule]. Do not present unverified AI output as checked expertise or another person's original work.

6. Report problems and hand over work

If information is shared incorrectly, an output is unsafe, or an agent acts unexpectedly, stop the affected activity and report it promptly to [incident contact or reporting route]. When someone leaves or changes role, hand over shared work, remove access through our leaver process and review subscriptions and integrations.

7. Keep the policy current

The policy owner maintains the approved tools register, answers questions and arranges a review by the date above, or sooner after a significant tool, use or risk change. Everyone using AI must read this policy and follow updates.

[Company name] AI acceptable use policy

Policy owner: [name or role and contact]
Approved by: [name or role] | Effective date: [date] | Next review: [date]
Approved tools and uses: [link to your AI tool register]

This policy applies to everyone using AI for [Company name] work. We use AI to support useful work while protecting company information and keeping people responsible for the results.

1. Approved tools and uses
Use only the tools, account plans and tasks in our approved AI tool register. Ask the policy owner before adopting a new tool, adding an integration or starting a paid subscription. Approval for one task or type of information does not cover every use.

2. Protect information
Never enter passwords, API keys or other access secrets. Use personal, client or confidential information only where the tool and task are explicitly approved for that type of information. Follow our existing privacy, security and client requirements. If unsure, use public or fictional material and ask before uploading.

3. Accounts and automated actions
Use company-approved work accounts and settings for company work. Do not use personal AI accounts. Get approval before giving an AI agent or integration access to company systems, and keep its access and actions within the agreed scope.

4. Check the result
The person using AI is responsible for the final work. Check important facts, sources, calculations, code and rights before relying on or sharing output. Consequential decisions about people need a separately approved process and accountable human review.

5. Be open about AI use
Disclose material AI assistance when it matters to the recipient, is required by a client, or falls under [company disclosure rule]. Do not present unverified AI output as checked expertise or another person's original work.

6. Report problems and hand over work
If information is shared incorrectly, an output is unsafe, or an agent acts unexpectedly, stop the affected activity and report it promptly to [incident contact or reporting route]. When someone leaves or changes role, hand over shared work, remove access through our leaver process and review subscriptions and integrations.

7. Keep the policy current
The policy owner maintains the approved tools register, answers questions and arranges a review by the date above, or sooner after a significant tool, use or risk change. Everyone using AI must read this policy and follow updates.

Make it your company's policy

Fill in the brackets. Add your company name, policy owner, approval and review dates, incident contact and disclosure rule.
List approved tools and uses. Name the account plan, permitted tasks and allowed information for each tool. Approval for writing public copy may differ from approval to handle client records.
Approve and share it. Align the wording with your existing company rules, tell the team where to find it and put the next review date in your calendar.

Template notes and sources

This is a starting policy for everyday company AI use. Adapt it to your contracts, data rules and sector requirements before approval. Get appropriate legal or data-protection review for regulated or sensitive uses.

Our wording draws on the security, human-control and governance principles in the UK Government AI Playbook and the NCSC secure AI design guidance, checked 8 October 2026. These sources concern government use and AI system development respectively; this template adapts the practical principles for staff using AI tools.

Common questions

How long should an AI policy be?

The downloadable template fits on one A4 page. Keep your core staff rules brief, and link to your approved tools register and any detailed procedures.

Do small companies need an AI policy?

A clear policy helps even a small team know which tools and information they may use, who checks the result and where to ask for help. The same starting template can be adapted by a larger company or individual department.