Quick Win
Your Red-Line Card

Take ten minutes and make your personal data-safety card — a sticky note or a phone note, your choice. Three sections: (1) my company's approved tools (look up the policy or send the ask-IT email today), (2) the Never List in your own shorthand, (3) the bulletin board test, written as a question.

Put it where you'll see it for two weeks. That's how long it takes for the check to become reflex — and once it's reflex, you can use AI faster than anyone around you, because you're the one who never has to stop and worry.

Prompt Pack: Working Safely

13.1The Genericizer Coach

I want to practice removing identifying details from prompts. Here's a made-up scenario with fake specifics: [INVENT ONE]. Show me the genericized version, then explain what you stripped and why the draft quality won't suffer.

13.2The Policy Decoder

Here is wording from a company technology policy (public/internal-general): [PASTE THE RELEVANT LINES]. Explain in plain English what this allows and forbids for AI chat tools, and list any questions it leaves unanswered that I should ask IT.

13.3The Ask-IT Draft

Draft a short, professional email to IT asking: which AI tools are approved for staff, what categories of information may be used with them, and whether guidance is written down anywhere. Tone: responsible and practical, not anxious. Under 90 words.

13.4The Ring Sorter

Quiz me: give me ten short office-task scenarios one at a time (invent them — generic office life). For each, I'll answer which is safest: company internal tool, enterprise built-in, public tool with genericized content, or no AI at all. Tell me if I'm right and why.

13.5The Team One-Pager

Create a one-page 'AI data safety basics' reference for office staff: the categories that never go in public tools, a plain-language version of the bulletin-board test, and the 'when unsure, ask' principle. Friendly, non-scary, formatted as a quick-reference sheet I can adapt to my company's policy.

13.6The Slip Protocol

Draft a calm, factual message reporting that I pasted information into a public AI tool that may not have been appropriate: what to include (what, when, which tool, what I've done since), tone professional and responsibility-taking without panic. Generic template with [BRACKETS].