Generative AI Acceptable Use Policy Template for Employees
Use this template as a starting point for your organisation's AI acceptable use policy. Replace bracketed text, adapt sections to your context and have it reviewed before adoption.
Short answer
A generative AI acceptable use policy template should include fifteen sections: purpose; scope; approved tools; data protection and confidentiality; acceptable uses; prohibited uses; accuracy and human review; transparency; intellectual property; fairness and bias; AI in decisions about employees; training and AI literacy; reporting concerns; breaches; and ownership and review. The template below provides draft wording for each, with bracketed fields to customise.
Key takeaways
- Customise bracketed fields and examples for your organisation.
- Link to a maintained list of approved tools rather than naming them in the policy itself.
- Include how the organisation uses AI on employees, not only how employees use AI.
- Have legal, data protection and, where relevant, employee representatives review it.
This is general information and a starting template, not legal advice. Adapt it to your organisation, jurisdictions, works council or union agreements and sector rules, and have it reviewed by qualified counsel.
AI acceptable use policy template
1. Purpose
This policy explains how [Organisation] employees, contractors and others working on our behalf may use artificial intelligence (AI) tools, including generative AI, in their work. It aims to help us use AI productively and responsibly while protecting our people, customers, data and reputation.
2. Scope
This policy applies to all use of AI tools for work purposes, on any device, including tools built into other software. It also sets out how [Organisation] uses AI in decisions and processes that affect employees.
3. Approved tools
You may use only the AI tools listed on [intranet page], which have been assessed for security and data protection. Each tool is approved for specific types of data. Using unapproved AI tools for work purposes, including personal accounts, is not permitted. To request a new tool, contact [owner].
4. Data protection and confidentiality
Never enter the following into an AI tool unless it is approved for that category of data: personal data about colleagues, candidates or customers; special category data such as health information; confidential business information; client information subject to confidentiality obligations; passwords or security details. Where possible, remove identifying details before using AI.
5. Acceptable uses
Examples include drafting and editing documents, summarising non-confidential information, brainstorming, research, coding assistance and data analysis, using approved tools and within the data rules above.
6. Prohibited uses
You must not use AI to: make or substantially determine decisions about individuals' employment, pay, promotion or discipline without authorised processes and human review; monitor colleagues outside approved processes; infer emotions of colleagues at work; create misleading, discriminatory, harassing or unlawful content; impersonate others; or circumvent security controls.
7. Accuracy and human review
AI can produce errors, outdated information and invented content. You are responsible for any work you produce with AI assistance. Check facts, figures, sources and legal or policy statements before use, and apply your professional judgement.
8. Transparency
Be open about significant AI assistance where it matters to the recipient, for example in external publications, client deliverables or where required by law. People interacting with AI systems we operate must be told they are interacting with AI where required.
9. Intellectual property
Do not enter third-party copyrighted material into AI tools unless permitted. Check that AI outputs do not reproduce protected content before publishing. Work created with AI assistance in the course of employment is subject to our usual intellectual property terms.
10. Fairness and bias
AI outputs can reflect bias. Review outputs, particularly those relating to people, for unfair or discriminatory content, and raise concerns with [owner].
11. AI in decisions about employees
Where [Organisation] uses AI in recruitment, performance, workforce planning or other people processes, we will inform affected people, keep meaningful human oversight of consequential decisions, test for fairness and provide routes to ask questions or challenge outcomes, in line with applicable law and consultation obligations.
12. Training and AI literacy
All staff who use AI tools must complete [required training]. Additional role-specific training applies to people who use or oversee AI systems in their work. This forms part of our measures to support AI literacy.
13. Reporting concerns
Report data incidents, harmful outputs, suspected misuse or concerns about AI systems to [contact] or through [reporting channel]. You will not be penalised for raising concerns in good faith.
14. Breaches
Breaches of this policy may result in action under our disciplinary procedure, depending on seriousness and intent.
15. Ownership and review
This policy is owned by [role] and reviewed at least every six months. Last reviewed: [date].
How to adapt the template
- Add sector-specific rules, for example for healthcare, financial services or legal work.
- Add examples relevant to your roles in sections 5 and 6.
- Align section 11 with local consultation duties and works council agreements.
- Link section 12 to your AI training program.
- Publish a short summary and FAQ alongside the full policy.
For a narrower ChatGPT-specific version, see ChatGPT use policy for employees.
Related guides
- AI Policy for Employees: Why You Need One and What It Should Cover
Why an employee AI policy is essential, what it covers and who owns it.
- What to Include in a Workplace AI Policy: A Section-by-Section Checklist
Twelve sections, the questions each must answer, and commonly missed items.
- ChatGPT Use Policy for Employees: Rules, Examples and a Short Template
Key rules, do and don't examples and a short template for ChatGPT and similar assistants at work.
- AI Training Program for Employees: A Ready-to-Adapt Curriculum
A seven-module AI training curriculum with audiences, durations, formats and assessment.
Frequently asked questions
Is there a free AI acceptable use policy template?
Yes. This page provides a free 15-section template with draft wording for each section, to adapt to your organisation and have reviewed by counsel.
What sections should an AI acceptable use policy have?
Purpose, scope, approved tools, data protection, acceptable uses, prohibited uses, accuracy and human review, transparency, intellectual property, fairness, AI in decisions about employees, training, reporting, breaches and review.
Should the policy name specific AI tools?
It is usually better to link to a maintained list of approved tools, because tools change faster than policies.
Does an AI policy need legal review?
Yes. Employment, data protection, AI regulation and consultation requirements vary by jurisdiction, so have the policy reviewed before adoption.