Appendix A. Everyday AI Safety Checklist

• Have I defined what I need the tool to do?
• Am I using an approved tool for this information?
• Have I removed passwords, personal identifiers, confidential records, and unnecessary details?
• Have I checked the output against reliable sources?
• Are quotations, citations, calculations, names, dates, and statistics verified?
• Could this output unfairly affect someone or mislead an audience?
• Have I disclosed AI assistance where required or where readers need to know?
• Is a qualified human reviewing the result if the stakes are high?
• Do I know how to report an error, impersonation, or security incident?
• Would I be comfortable explaining how I used AI and why I trusted the result?