# Workplace AI Disclosure Decision Guide

Use with [Workplace AI Disclosure: When You Actually Have to Say So](/articles/workplace-ai-disclosure-when-required).

## Check each trigger in order

**1. Does my company's AI policy name an explicit disclosure requirement for this kind of work?**

- [ ] Yes — follow the policy exactly. Stop here.
- [ ] No / unsure — checked using the [workplace AI policy read card](/downloads/articles/workplace-ai-policy-read-card.md) and continue.

**2. Does a client contract or statement of work specify disclosure for this deliverable?**

- [ ] Yes — follow the contract. Confirmed with the contract/account owner if I'm not sure it applies.
- [ ] No / unsure — continue.

**3. Does my work build, configure, or operate something that interacts directly with customers or the public using AI (chatbot, voice agent, automated response)?**

- [ ] Yes — this is a compliance question, not an individual judgment call. Under the EU AI Act the duties split by role (providers under Article 50(1); deployers under 50(3)-(4)), which is for compliance/legal to determine, not you. Escalated to compliance/legal/product: ___________ (name/date).
- [ ] No — continue.

**4. Does a formal review, audit, or intake form ask directly whether AI was used?**

- [ ] Yes — answered honestly and specifically.
- [ ] No — continue.

**5. None of the above apply.**

- [ ] Use the general disclosure norm: would the person relying on this work want to know, and would they feel misled without it? See [keeping your name on AI-assisted work](/articles/keep-your-name-on-ai-assisted-work).

## If the answer was unclear

- [ ] Asked a named person, in writing.
- [ ] Defaulted to more disclosure, not less, while waiting for an answer.
- [ ] Kept a record of who I asked and when.