Guardrails

Protecting data and people while you use AI.

Tools

• Use only approved, secure AI tools for your organization
• Confirm what licenses and policies your employer provides before adding new apps
• Check whether AI is already built into tools you use daily (Copilot in M365, etc.)
• Avoid unverified or “free” tools when content is sensitive or internal

Safe use

• Never paste or upload sensitive or confidential details into unapproved tools
• Always double-check AI output before sharing — models make mistakes
• Watch for bias, especially in hiring, performance, or high-stakes decisions
• If using open-source or self-hosted AI, involve IT/security — don’t configure it alone

Best practices

• Read your organization’s privacy and AI use policies first
• Use approved platforms for anything that identifies people or projects
• Give context: your role, audience, and the format you need
• Review outputs the way you would a junior teammate’s draft — carefully