Storytelling with AI: Do you need an “AI Use Policy”?

Creating an AI Use Policy in real time.

Most people come into StoryCorp’s Storytelling with AI training with similar questions. They’re looking for ways to get better at using AI, unlock some of their own creativity, save time, and learn new and dynamic ways to work with their material. But once people get better at using AI, there are two less-obvious questions that tend to pop up:

When you’re telling stories with AI, should you have an explicit policy in place? And should that policy be personal, public, or both?

It’s up to the individual (or the individual’s employer!) to answer these questions for themselves. But if you’ve decided that you want to formalize your approach, here are some things to keep in mind

What’s an “AI Use Policy” anyway?

At its best, it’s a short, plain-language story about how you work with AI:

• What’s allowed and encouraged

• What’s not allowed

• How you protect people’s data and trust

• What quality checks and approvals are needed

• How you learn and improve over time

Think of it as an agreement between humans and their new digital collaborator.

It should be short (one or two pages max), actionable (clear do’s and don’ts with examples), aligned with your values (especially if you do public-facing storytelling), and living (easy to update as tools, risks, and opportunities evolve).

A young professional takes a coffee break while thinking hard about whether he really needs AI Use Policy.

When you probably don’t need a formal policy (yet)

You might not need a formal policy if most of the following are true:

• You’re a small team (e.g., a few people) experimenting informally.

• You’re only using AI for low-risk tasks: brainstorming, idea generation, rough first drafts.

• You rarely handle sensitive or confidential information.

• Nothing you create with AI is published externally without human review.

• You’re not in a highly regulated space (health, finance, government, children’s data, etc.).

In that situation, a short “AI house rules” paragraph in your team handbook (or on a Post-It Note by your desk) might be enough for now.

If you often receive flowers from robots, you should probably have an AI Use Policy.

When you should consider a formal policy

On the other hand, you should consider a clear AI Use Policy if any of these ring true:

• You publish stories that affect public trust or regulatory decisions.

• Your team works with confidential data

• You’re already using AI tools and starting to see uneven practices.

• You’ve had at least one “uh-oh” moment: a near miss with privacy, accuracy, or tone.

• Leadership is asking, “Are we sure this is safe?” or “What’s our official position?”

If you’re telling important stories with AI in the mix, you want a shared guardrail that keeps creativity high and risk low.

Where to begin?

If you’re curious about what the kernel of an AI Use Policy could look like, check out the quiz below. It’s not perfect, but it covers many aspects of our relationship to AI, and will give you a starting point, as well as some food for thought about how you should approach AI. It also happens to be (full disclosure!) developed with AI—the question categories, the questions themselves, and even the code that computes your answers and gives you the output. In today’s world, would we expect anything less than using AI to help us think about how to use AI?

Your AI Use Style Quiz

Answer honestly. At the end, you'll get a short "AI Use Policy" you can copy into a syllabus, team charter, portfolio, or bio. Nothing here is stored or sent — scoring happens entirely in your browser.

1. Which statement sounds most like you when it comes to AI and authorship?

I want the words to be mine. AI can nudge commas or clarity, but it shouldn’t write for me. AI can help with phrasing, structure, and clarity — but the ideas and final voice are mine. AI can draft from scratch. My job is to refine, fact-check, and take responsibility. If it's routine work, I'm fine letting AI write it end-to-end under my name.

2. When do you think AI use should be disclosed?

Anytime AI meaningfully contributes ideas, structure, or wording. Transparency builds trust. I'll disclose if AI shaped the thinking or language in a visible way — especially in public / professional work. I’ll only mention AI if someone asks directly. I don’t really rely on AI for original content, so disclosure usually doesn’t apply.

3. What is writing, to you?

Writing is thinking. Outsourcing that core thinking to AI feels wrong. Writing is thinking and communicating. I’ll keep the thinking, and let AI help with structure and polish. Writing is delivering a usable asset on time. AI is a valid first drafter that I’ll then refine. Writing is output. If AI gets me that output faster, that’s success.

4. Who is responsible for accuracy, facts, and claims in AI-generated text?

If AI wrote it and I use it, I’m 100% responsible for verifying every claim before it goes out. AI is a brainstorm partner, not a source of truth. I double-check with real sources before I trust anything. If it sounds plausible and saves time, I’m comfortable using it with minimal checking. I wouldn’t rely on AI for factual or high-stakes claims at all.

5. How do you feel about relying on AI long-term?

I want to stay fully capable without AI. I treat AI as optional, not a crutch. I’ll lean on AI for speed, but I actively keep my own skills sharp. If AI vanished tomorrow, I’d struggle — and I accept that tradeoff for efficiency today. I’m already outsourcing most of the heavy lifting, and that’s intentional.

6. How important is it that the final text “sounds like you”?

Very. My judgment, taste, and lived experience have to be on the page. I care about clarity and usefulness more than sounding uniquely like me. If it solves the task, I don’t need to touch it much. If AI makes everyone sound the same, that’s a problem. I’d rather sound human, even if it’s slower.

7. What about personal or values-driven writing (a cover letter, a speech “from the heart,” an apology)?

Those need to come directly from me, not AI. I’ll draft those myself, but I’m fine asking AI to tighten tone or clarity. AI can do a first pass, then I rewrite until it feels true. I’m comfortable letting AI generate even personal or persuasive messages if I’m signing my name.

8. How comfortable are you telling your audience (boss, client, reader, public) that AI helped?

Very comfortable. I think saying “AI-assisted” builds trust. I’ll say it when it shaped the thinking or wording in a meaningful way. I’d rather not draw attention to AI unless someone directly asks. This mostly doesn’t apply, because I keep AI’s role minimal in substantive writing.

9. How do you treat confidential / sensitive / internal information (client data, HR issues, health info, legal strategy, etc.)?

I will never paste confidential or regulated info into AI unless the tool is explicitly approved for that use. I’ll sanitize or anonymize sensitive material first. Protecting people comes before convenience. I’ll paste internal material if the tool looks trustworthy and I’m in a rush. If a tool markets itself as “secure,” I’m comfortable dropping in drafts, client names, etc.

10. Which priority best reflects your personal AI ethic?

First, do no harm. Privacy, fairness, and avoiding damage to real people outrank speed. Integrity of voice and craft matters most. I don’t want to lose what makes my work mine. Responsible openness matters. I want clear norms and I’d rather err on the side of transparency. Quality and accuracy matter most. I’ll take ownership of whatever I publish, even if AI drafted it. Efficiency matters most. If AI saves hours, we should be using it.

See My AI Use Policy

Your answers never leave this page. Scoring is done with simple JavaScript in your browser.

Your AI Use Style: