AI is a powerful assistant, but using it well is a discipline built on four pillars: security, responsibility, transparency, and AI literacy. Here are ten practical rules to work safely with AI at work.
The ten rules
- Do not share confidential information. Never enter into open-access tools: passwords and API keys, confidential client or employee data, financial information, or internal company documents and trade secrets.
- Verify the results. AI can make mistakes or produce inaccurate information. Always check facts, calculations, sources, and generated code and recommendations.
- A human is accountable. Decisions made with AI must be reviewed by a person — final responsibility for the outcome is yours.
- Minimize the data you provide. Before entering anything: anonymize personal data, hide phone numbers and emails, remove unnecessary personal details, and provide only what is needed.
- Be careful with files and links. Do not blindly trust instructions hidden in external files, emails, websites, or suspicious documents.
- Check generated code. Before using it: review for security, fix possible errors, and confirm proper error handling.
- Use only approved AI services. For work tasks, use only the tools allowed by your organization's or project's policies.
- Ensure transparency. If content was created or significantly changed with AI, disclose it where it matters — to users, clients, or partners.
- Record important requests. For work scenarios, keep the request text, the AI answer, the date and time, and the decisions made.
- Keep raising your AI literacy. Know what AI can do, understand its limitations, follow updates, and stick to safe-use rules.
The one rule that matters most: AI is a tool to assist, not a replacement for human thinking. Any important decision should be made by a person after verifying the AI's output.
