ai-usage-policy-generator
Produces a comprehensive, ready-to-review AI acceptable use policy document customized to your industry, regulatory requirements, and risk appetite. Covers data classification, approved tools, prohibi
Files1
View on GitHubSKILL.mdskills/ai-usage-policy-generator/SKILL.md
GitHubAI Usage Policy Generator
Generate a tailored AI acceptable use policy for your organization based on industry, size, and risk tolerance.
Integrations: Google Drive
When to Use
- The user needs to create an AI acceptable use policy
- An organization is adopting AI tools and needs formal guidelines
- The user mentions "AI policy", "acceptable use", or "usage guidelines"
- Legal or compliance teams request documentation for AI tool usage
Steps
Step 1: Gather Policy Inputs
Ask the user for:
- Industry (finance, healthcare, tech, education, government, etc.)
- Organization size (startup, mid-market, enterprise)
- Risk tolerance (conservative, moderate, progressive)
- Regulatory requirements (SOC2, HIPAA, GDPR, FedRAMP, PCI-DSS, none)
- AI tool categories in scope (code generation, chat assistants, image generation, data analysis)
Step 2: Generate Policy Sections
Create a comprehensive policy document with these sections:
| Section | Content |
|---|---|
| Purpose & Scope | Why the policy exists, who it applies to, what tools it covers |
| Approved Tools | Categories of approved AI tools, approval process for new tools |
| Data Classification | What data can/cannot be shared with AI tools by classification level |
| Prohibited Uses | Specific use cases that are not allowed (e.g., PII in prompts, code with trade secrets) |
| Security Requirements | Authentication, encryption, data retention, logging requirements |
| Intellectual Property | Ownership of AI-generated content, license compliance, attribution |
| Compliance & Audit | How usage is monitored, audit frequency, reporting requirements |
| Incident Response | What to do if data is accidentally shared, escalation path |
| Training Requirements | Mandatory training before AI tool access, ongoing education |
| Policy Review Cadence | How often the policy is reviewed and updated |
Step 3: Customize by Risk Profile
Adjust the policy strictness based on risk tolerance:
- Conservative: Allowlist-only tools, mandatory review of all AI outputs, no external AI APIs, quarterly audits
- Moderate: Approved tool catalog with self-service request process, spot-check reviews, semi-annual audits
- Progressive: Default-allow with blocklist, trust-but-verify approach, annual audits with continuous monitoring
Step 4: Generate Google Drive Document
Create a well-formatted Google Doc with:
- Company logo placeholder and document metadata (version, date, owner)
- Table of contents with hyperlinks
- Each section with clear headings and actionable language
- Appendix with tool-specific guidelines
- Signature/acknowledgment page template
Step 5: Provide Rollout Guidance
Recommend a rollout plan:
- Legal review — Have legal/compliance review the draft
- Stakeholder feedback — Share with engineering leads and security team
- Employee communication — Draft an announcement email
- Training — Create or recommend training materials
- Acknowledgment — Process for employees to sign/acknowledge the policy