Build a Team AI Policy

Why you need this

Teams using AI need clear guidelines. Without a policy, people either avoid AI entirely (missing the benefits) or use it recklessly (pasting confidential data into public tools). A simple one-page document prevents both problems.

Step 1. Define scope

• •List which AI tools are approved (e.g., ChatGPT, Claude, Copilot).
• •List which tasks can use AI (drafting, research, coding) and which cannot (legal advice, medical advice, anything requiring professional judgement).
• •Keep it short — one page max.

Step 2. Set data handling rules

• •No customer PII in public AI tools (don't paste names, emails, addresses, or financial data into ChatGPT, Claude, or Gemini).
• •No confidential company data (financials, strategy, client lists).
• •For sensitive work, use approved tools only or local AI (like Ollama).

Step 3. Set disclosure requirements

When must employees disclose AI use?

• •Client-facing work: always disclose.
• •Internal documents: not required but recommended.
• •Code: add a comment noting AI assistance.
• •Be clear about when AI output is presented as human work vs. AI-assisted.

Step 4. Set quality standards

• •AI output must be reviewed by a human before use.
• •Fact-check all claims and statistics.
• •Verify code runs correctly before deploying.
• •Spell-check and tone-check written content.
• •The employee is responsible for the final output, not the AI.

Step 5. Template policy

Provide a copy-paste template for your team to adapt.

Policy Template

Step 6. Implementation

1. 1.Review the policy with your team in a short meeting.
2. 2.Get sign-off from key people.
3. 3.Train staff on approved tools and prompt patterns.
4. 4.Review quarterly and update as tools evolve.
5. 5.A policy that isn't communicated doesn't exist.