The short answer
Using AI to help write a page doesn’t rule it out of citations on its own. Google’s guidance allows AI assistance, and OpenAI’s crawler documentation says nothing about how a page was written. The bigger risks are errors, generic advice, and large volumes of unreviewed pages.
No writing method can guarantee a citation, whether a person, an AI tool, or both produced the page.
What does Google say about AI-written content?
Google’s guidance on generative AI content says AI can help with research and with adding structure to original content. The warning applies when tools create many pages without adding value for users, which may breach its scaled content abuse policy.
Accuracy is part of the same picture. In October 2026, Google added a warning that AI output can contain errors, and it now says manual fact-checking is critical before publishing. That review also covers titles, meta descriptions, structured data, and image alt text.
Google’s guide to generative AI in Search goes further on quality. It warns against commodity content that restates what is already available or could easily be produced by a generative AI model.
Picture an accountant’s blog. “Seven tax tips for small businesses” could come from almost anywhere. A piece on the BAS mistakes the firm keeps seeing in first-year clients draws on experience other sites can’t copy.
Do ChatGPT and Perplexity treat AI-written pages differently?
We couldn’t find a published rule from OpenAI or Perplexity saying that AI-written pages are rewarded or penalised. Their public crawler guides deal with access, and our guide to AI crawler settings explains those controls.
Without a published rule, treat confident claims about how either tool scores AI writing with caution. Studies do exist, but many come from companies that sell SEO or AI detection software, and they rely on classifying pages as human or AI. That leads to the next problem.
Can AI detectors tell you whether a page is at risk?
Not reliably. OpenAI withdrew its own AI text classifier in July 2023 because of its low accuracy. Australia’s higher education regulator, TEQSA, has said testing keeps showing these tools are unreliable.
TEQSA’s work relates to education, but the limit applies to the tools themselves. A detector score can’t show whether your page is accurate, useful or worth citing, so your editing time is better spent on those qualities.
Which AI workflows carry more risk?
This table reflects our judgement based on Google’s guidance. It isn’t a measured risk score.
| How AI is used | Risk | Why | What to do |
| Research and outlining | Lower | A person still decides what is published | Check every source yourself |
| Drafting from expert notes | Lower | The ideas come from real experience | Have the expert approve the final page |
| Generic prompt with a light edit | Higher | The page may repeat what is already online | Add first-hand insight or don’t publish |
| Bulk pages for each suburb or service | Highest | Volume without value may breach Google’s policy | Build pages only where you have something distinct to offer |
| Regulated advice without expert review | Highest | Errors can create legal and compliance problems | Use a qualified reviewer and keep a record |
Repetition can hurt even when every page is accurate. Our article on similar articles and content overlap explains why five versions of one answer rarely help.
How should you review AI-assisted content?
- Trace each claim: Open the original source and confirm the date, wording, and context.
- Remove what anyone could say: Add a real example, process, or decision your business can stand behind.
- Compare it with your website: Services, prices, timelines, and advice should match what you’ve already published.
- Localise it properly: Use Australian English and references your customers will recognise.
- Check the metadata: Review titles, meta descriptions, structured data and image alt text as well as the article.
- Give someone responsibility: A named person should approve the final page.
At Myoho Marketing, our content review checks for inconsistencies, Australian English, and any conflict with our existing blogs.
Should you tell readers you used AI?
There’s no single rule for every page. Google’s guidance on AI content says sharing how content was created can give readers useful context. The National AI Centre’s voluntary guide, Being clear about AI-generated content, says transparency can build trust and reduce regulatory risk. The guide itself states that it was produced with AI assistance while the team kept full editorial control.
The Centre’s AI transparency statement gives more detail for higher-impact content, and that’s a sensible model to follow. Disclosure won’t fix a misleading claim, so your review process still matters. Regulated businesses should also read our guide to AI-assisted content in regulated industries before publishing.
Frequently asked questions
Does Google penalise AI content?
Google’s guidance doesn’t treat AI use alone as a breach. Its stated concern is creating many pages without adding value for users.
Should we stop using AI for blog writing?
No need, if people stay responsible for the facts, ideas, and final approval.
Do AI tools prefer human-written pages?
We haven’t found a published rule from OpenAI or Perplexity that says so.
Should we run old posts through an AI detector?
Reviewing them for accuracy, overlap, and usefulness will tell you far more.
Build content worth citing
Want a clear view of which pages are helping your AI visibility and which ones need work? Our Generative Engine Optimisation services cover research, content and visibility tracking, and you can book a free consultation to talk through your site.
This article is general information, not legal advice.







