Does Google penalise AI-written content?

You want the blog and the service pages your business never has time to write, AI makes that suddenly affordable, and then someone tells you Google will bury the site for it. So the content does not get written, again. The fear is out of date, and Google itself is the source for that.

Google's published guidance on AI-generated content says it plainly: appropriate use of AI or automation is not against its guidelines, and its systems reward quality content "however it is produced". What Google does penalise is using automation to churn out pages whose primary purpose is manipulating rankings, the spam policy known as scaled content abuse. Note what that policy targets: the churning, not the tool. A thousand interchangeable pages written by underpaid humans get the same treatment.

The independent data agrees. Semrush's study of AI content and rankings found AI-assisted and human-written content reaching the top ten at nearly identical rates, 57% against 58%. There is no AI detector deciding your fate; there is a quality bar, and most content of every origin misses it.

What this means in practice

The question to retire is "will Google punish AI content". The question that decides your rankings is "would anyone choose to read this page". AI-written sludge fails that test, fifteen hundred words of correct, sourceless, interchangeable nothing that could sit on any competitor's site with the logo swapped. No penalty needed; it simply loses to better pages.

What wins is the material only you could publish, with AI doing the labour: your prices, your timescales, the questions your customers ask in week one, the job that went wrong and what it taught you, photographs of your work. A plumber's page on what a Stockport boiler swap costs, with real numbers and real caveats, beats a thousand words of generic "boiler replacement considerations" every day of the week. The draft can come from a model; the substance has to come from your business. That substance is also what Google's quality framework (experience, expertise, authority, trust) is trying to measure.

So the working method is: AI drafts, you inject the specifics and the opinions, every claim gets a source or gets cut, and a human reads it before it ships. Publish at whatever pace keeps that bar, and stop worrying about the byline. Most of what I publish starts as an AI draft, and it ranks because every number is checked and sourced, not because anyone is hiding how it was made.

Two related notes. The same logic now extends past Google: ChatGPT and AI Overviews cite pages that are specific and well-sourced too, and we cover that in how to get your business cited by ChatGPT. And remember the model will cheerfully invent statistics for your article, which is your problem, not Google's: see why ChatGPT makes things up before you publish anything with a number in it.