At OpenOffice.org.nz we want to be open about how our content is made. We use artificial intelligence (AI) as one tool in our editorial process, but every guide is reviewed, fact-checked and edited by a person before it is published. This page explains how that works.
Last updated: October 2026.
How we use AI
We use the AI tool Claude (developed by Anthropic, claude.ai) to make parts of our work more efficient. Typical uses include gathering and summarising research, proposing a structure for an article, producing a first draft and suggesting clearer wording. AI tools may also be used to help create illustrations for our articles.
AI is always an aid and never the final word. No text is published in the form it leaves the AI tool.
We do not enter personal information into AI tools. Messages that readers send us are not used as input for AI drafting.
Our human editing
After the drafting stage, an editor takes over. This is where the most important work happens. Among other things, we:
- Fact-check details against reliable primary sources, such as the official documentation, release notes and security notices published by the Apache OpenOffice project, The Document Foundation, Microsoft and Google.
- Check instructions – menu names, file formats, system requirements and steps – against the developer’s own guidance, because AI tools can describe features that have moved, been renamed or never existed.
- Correct or remove anything that is inaccurate, outdated or misleading.
- Rework the language so that the text is clear, accurate and easy to follow, including for beginners.
- Add context an AI tool cannot supply on its own, such as what a change means for a New Zealand household, school or small business.
- Make sure the content is original and not copied from other websites.
The aim is that you always get content that is correct, useful and written for people rather than for search engines.
Why we are open about it
We believe in transparency. Telling you how our articles are produced is a matter of trust: you should be able to rely on what you read here and know how it came about.
We also follow Google’s guidance on helpful, reliable content. Under that guidance, what matters is that content is accurate and made to help people, whether or not AI was involved in producing it. That is the standard we work to.
Fact-checking and responsibility
Everything we publish is covered by editorial responsibility. A responsible editor reviews the material, and we update our pages as the facts change – for example when a new software version is released, a subscription plan is restructured or a product reaches the end of its support. You can read more about who we are on our About us page.
Software changes quickly, and AI tools can be out of date or simply wrong. That is why details that affect your decisions, such as prices, licence terms and security status, are checked against the supplier’s own information before publication.
Our content is information, not professional advice
Our articles are informative and descriptive. They are not personalised IT, licensing, security or legal advice, and they are not a recommendation to choose a particular product. Before you install software, buy a subscription or move important documents, confirm the details with the software’s developer or a qualified adviser. Our terms of use explain this in more detail.
Found an error?
We work hard to get things right, but mistakes can still slip through. If you spot something that is wrong or out of date, we would be grateful if you let us know through our contact page, ideally with a link to the page in question. We will correct it as soon as we can.