# Writing an AI disclosure policy your team will actually follow

> Vague policies get ignored. A clear one fits on a page, names the tools and says who checks the work. A starting point for newsrooms, agencies and content teams.

Canonical page: https://winstonai.ai/blog/writing-an-ai-disclosure-policy

By [Priya Natarajan](https://winstonai.ai/authors/priya-natarajan), Content strategist. Published 2026-08-20. Guides, 5 min read.

Most organizations now have some kind of AI policy. Few are followed, usually because they're either too vague to act on or too strict to be realistic. A good disclosure policy is short, specific and easy to check.

## What a useful policy covers

- Which tools are approved, and for which tasks.
- What must be disclosed to editors, clients or readers, and where.
- Who reviews AI-assisted work before it's published.
- How work is checked, and what happens when something slips through.

> Disclose the tool, the task and who checked the result.

## Keep disclosure simple

Readers don't need a technical breakdown. A single line such as "Research and outline assisted by AI; written and fact-checked by our editors" is honest and clear. Internally, ask contributors to note AI use in the same place they track sources.

## Check, don't just trust

A policy without verification turns into an honor system. Spot-check submissions with a detector, especially from new contributors, and treat high scores as a reason to ask questions. The aim is consistency, not surveillance.

## About the author

Priya Natarajan, Content strategist. Priya writes about search, publishing and editorial standards for teams that use generative AI.