Flink Ai Use Statement

---
name: flink-ai-use-statement
description: Writes your AI Use Statement, how you actually use Claude in your writing, in your own words as one line and one paragraph, checked against policy, audience, context, expectations and degree of use, and never hiding AI use. Use for "run flink-ai-use-statement", "did AI write this", "how do I disclose AI use", "should I say I use Claude", "AI disclosure for my profile", "honest answer about AI in my posts", "write my AI use line", part of the Claude Playbook for Authentic LinkedIn Presence Pack by Polar Bear.
---

# AI Use Statement

## When To Use
Someone asks "did AI write this?" under your post, in a DM or on a sales call, and you want an honest answer ready instead of an awkward one. Use this once, keep the result on your profile or in your notes, and revisit it when the way you work changes.

## When Not To Use
If you want to argue what you believe about AI in your field, that is a belief, and it belongs in Point of View Statement or Opinion Post. If an employer, client contract or regulator sets disclosure terms for you, start from those terms and check with a qualified adviser; this skill writes your words, not a compliance notice.

## Inputs
- A plain description of how you write today: what you do, what Claude does, what nobody else touches.
- Any rule you already work under (an employer or client policy, a platform term you know of).
If you have none of this, I start from three questions about your last post and mark the output as a first draft.

## Approach
The check is NN/G's PACED framework (Policy, Audience, Context, Expectations, Degree of AI use), published by Taylor Dykes in September 2026. It sits beside the transparency dilemma research (Schilke and Reimann, 2025): across experiments, people who disclosed AI use were trusted less, and people whose AI use was exposed by someone else were trusted least of all. That cost is real and the statement does not pretend otherwise; it simply never chooses to hide. The failure this prevents is the vague line ("AI-assisted") that answers nothing and reads as if something is being kept back.

## Workflow
1. Ask three questions: walk me through your last post from idea to publish; which parts would you never hand over; who asked, or might ask, and why it matters to them.
2. Describe actual use first, step by step: where the idea came from, what Claude did (interviewed you, found ideas in your notes, checked voice and claims, suggested cuts) and what only you did (the ideas, the final words, the decision to post). Keep it factual and specific.
3. PACED pass. Policy: any rule you work under, and any legal disclosure duty (check with a qualified adviser). Audience: who reads you and what they assume. Context: is sincerity or originality central to this kind of post. Expectations: would a reader feel misled if they learned the process later. Degree: how much of the final text is Claude's.
4. Draft one line for your profile and one paragraph for when someone asks, in your words from the voice guide if you have one. Both state the use plainly; neither apologises or oversells.
5. Test each against two readers: a client who prefers no AI at all, and a client who uses it daily. Would either feel misled? If yes, the statement is not yet accurate.
6. Set a review trigger: the statement is updated when the way you use Claude changes, not on a calendar.

## Output Format
```markdown
# AI Use Statement
## How I actually write
| Step | What I do | What Claude does |
|---|---|---|
| [idea, draft, check, edit, post] | [your part] | [Claude's part or none] |
## PACED check
| Lens | What applies to me |
|---|---|
| Policy | [rules; legal points: check with a qualified adviser] |
| Audience | [who reads and what they expect] |
| Context | [where sincerity or originality matters most] |
| Expectations | [would anyone feel misled later] |
| Degree of use | [how much of the final text is Claude's] |
## One line
[For the profile or a reply]
## One paragraph
[For when someone asks]
## Decision
[You approve both versions, decide where the line appears, and name the change that will trigger a review, by [date].]
```

## Done When
- The description of use matches what you actually do, step by step.
- All five PACED lenses are answered.
- The line and the paragraph are in your words and say what Claude does.
- A review trigger is set.

## Quality Bar
- Accurate over reassuring: no "lightly assisted" unless that is true.
- No claim about how readers will react; the research is described, not turned into a promise.
- Never sells against AI and never apologises for using it; legal disclosure points end with "check with a qualified adviser".
- An honest statement in your words; the pack never hides AI use.

## Next
Run flink-project-story-post (Project Story Post) to write the first post.

## About the makers

This pack is made by Polar Bear, a consultancy built by ex-McKinsey founders with a dream to make AI work for People, not instead of them. We help our clients build people systems and AI-first ways of working, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).