ConeheadAI · leftover lesson
SNK3-097 · Curriculum How-to
Image AI at work: brand-safe prompt + human publish gate
L0–L2 · Curious–Capable (bootstrap — system rebalances) · ~3m
Look for this insight
Treat image/voice generators like any workplace tool: clear brief, no client faces or confidential UI, and a human gate before the asset ships.
Lesson hook
Image tools like Copilot Create, ChatGPT images, Gemini, and Midjourney are fast — and they'll leak context right along with you if your prompt includes client names, unreleased UI, or real employee photos. The workplace pattern: write a brand-safe brief (audience, style, must-nots), generate the image, then keep publish human-owned. Same rule as chat: capability isn't permission to ship.
Do this now
Open your notes (or your image tool, if you have one). Write a 5-line image brief for one real work asset: (1) audience (2) style (3) must include (4) must NOT include (PII/client/brand-unsafe) (5) who approves publish. You're done when you've circled the one line you usually skip.
Completion check: writes_image_brief_must_nots
Staff review Sign in