A ready-to-edit generative AI policy template that governs how employees create content with AI. Covers ownership and copyright risk of generated work, disclosure of AI-generated customer-facing content, brand-voice and accuracy standards, image and media generation rights, AI-generated code review, and prohibited generation like deepfakes and impersonation.
Who needs this
Any small business whose team is generating customer-facing text, marketing images, or product code with tools like ChatGPT, Midjourney, DALL-E, or GitHub Copilot. If AI-written blog posts, AI-generated images, or AI-suggested code are already going out under your brand, you have copyright, accuracy, and disclosure exposure that a general acceptable-use policy does not fully cover.
What's inside
- Scope and definitions specific to generative AI (text, image, media, and code)
- Ownership and copyright-risk rules for AI-generated work
- Disclosure requirements for AI-generated customer-facing content
- Brand-voice and factual-accuracy standards for generated text
- Image and media generation rights, including likeness and trademark risk
- AI-generated code review, license, and security obligations
- A prohibited-generation list (impersonation, deepfakes, unreviewed legal/medical/financial claims)
- An approved-generation-use-cases table you fill in
- An employee acknowledgment and signature block
Preview
Generative Artificial Intelligence Policy
Effective date: [DATE] · Version 1.0 · Owner: [POLICY OWNER / ROLE]
1. Scope and Definitions
"Generative AI" means any tool that produces new text, images, audio, video, or code from a prompt, including but not limited to ChatGPT, Claude, Google Gemini, Midjourney, DALL-E, Adobe Firefly, and GitHub Copilot.
"Generated Work" means any output created with the help of a generative AI tool, whether used as-is, edited, or used as a starting point.
"Customer-Facing Content" means any Generated Work that a client, customer, or the public will see or receive, including marketing copy, blog posts, social media, product descriptions, images, emails, and code that ships in a product.
This Policy applies whether the generative AI tool is accessed on a Company device or a personal device, and whether or not the Company pays for the tool.
2. Ownership and Copyright Risk of Generated Work
Generated Work that an employee creates for the Company belongs to the Company on the same terms as any other work product. However, the Company understands that purely AI-generated output may not qualify for copyright protection in some jurisdictions, and that the Company cannot own what it cannot copyright.
To protect the Company's ownership position, employees must add meaningful human authorship to Generated Work before it is published or delivered — editing, arranging, and revising it rather than publishing raw AI output. Employees must not assume the Company owns exclusive rights in unedited AI output.
Employees must not submit third-party copyrighted material, competitor content, or another person's work into a generative AI tool as a prompt to reproduce or closely imitate it, because the resulting output may infringe that party's rights.
The full template continues with 9 sections. Grab the editable Word file using the form, then customize the bracketed [PLACEHOLDERS] for your business.
How to use it
- Download the editable Word file and add your company name and effective date.
- Fill in the Approved Generation Use Cases table with the content types your team is allowed to create with AI.
- Adjust the disclosure rules to match your client contracts and industry (add regulated-claim rules for finance, health, or legal work).
- Circulate to everyone who creates content, code, or media, collect signed acknowledgments, and review every 6–12 months.
Frequently asked questions
- An AI acceptable use policy covers all AI tools and the rules of use — which tools are approved, what data may never be entered, and who is responsible for output. A generative AI policy goes deeper on the act of creating content: who owns AI-generated work, the copyright risk, when you must disclose AI-generated customer-facing content, and specific rules for generated text, images, and code. Most businesses adopt both because they cover different risks.
- Not automatically. In several jurisdictions, purely AI-generated output may not qualify for copyright protection, which means you may not be able to claim exclusive rights in raw AI output. The practical fix, built into this template, is to require meaningful human editing and authorship before AI-generated work is published or delivered, so the finished work is genuinely yours.
- It depends on your contracts, the platforms you publish on, and the type of content. This template requires disclosure or client approval where a contract demands human-created work, where an AI image would be mistaken for a real photo, and where a platform requires AI media to be labeled. Testimonials and reviews must never be AI-generated and presented as real. When in doubt, disclose.
- It becomes an enforceable internal policy once you adapt it to your business, adopt it formally, and have employees acknowledge it in writing. Because copyright, likeness, and disclosure law varies by location and is still evolving for AI, have counsel review it before rollout. This template is a starting point, not legal advice.
This template is provided by Layer3 Labs for general informational purposes only and does not constitute legal advice. Copyright, likeness, and disclosure laws vary by jurisdiction and industry, and the law on AI-generated work is evolving. Have this policy reviewed by qualified legal counsel before adopting it.