Is AI-written content against Google’s rules? The policy says something more specific.

Two people have told you opposite things. One says Google penalises AI content and you will be de-indexed for using it. The other says Google does not care and has a package of forty posts a month ready to go.

The policy is public, it is short, and it does not say either of those things.

What the rule actually is

Google's spam policies name something called scaled content abuse, defined as many pages being generated for the primary purpose of manipulating search rankings and not helping users.

Read that again and notice what is missing. There is no mention of who or what produced the words. The policy's examples put generative AI tools alongside scraping and automated transformation of other people's work, which tells you the method is not the thing being judged.

A human can write scaled content abuse by hand. It would just take longer.

The whole rule hangs on one word

Primarily. Pages made primarily to manipulate rankings rather than to help anyone.

That word turns a technical question into an editorial one, and it is why "was this written by AI" is not the question to be arguing about. A researched, sourced, genuinely useful article that an AI helped draft passes the test. A hundred spun variations of the same page aimed at long-tail phrases fails it, and would fail it if a person typed every one.

Google's own guidance elsewhere makes the target explicit: using automation primarily to manipulate rankings is a spam violation. Not using automation. Using it for that.

The question nobody selling you content mentions

In its guidance on creating helpful content, Google publishes a self-assessment built around who created the content, how it was produced, and why it exists. Under "how" sits a question worth quoting in full: "Is the use of automation, including AI-generation, self-evident to visitors through disclosures or in other ways?"

It goes further and says AI or automation disclosures are useful for content where someone might think "How was this created?"

That is softer than a requirement and firmer than nothing, and it is absent from every AI content pitch we have seen. If a service is producing work under your name and your face, the question of whether a reader would want to know how it was made is one you should answer deliberately rather than discover later.

For the record: we use AI in our own writing process, including on this blog. The research, the sourcing, the arguments and the judgement calls are ours, every factual claim gets checked against a primary source before it goes up, and we would rather say that plainly than have you wonder.

The line aimed straight at the product being sold

The same self-assessment asks whether content is primarily made to attract visits from search engines. Then it asks something sharper: did you decide to enter some niche topic area without any real expertise, but instead mainly because you thought you would get search traffic?

Sit that next to the offer of forty posts a month covering every question in your industry. A plumbing company does have real expertise, and a post about what actually fails in a water heater is worth publishing. The same company publishing on commercial HVAC procurement because a keyword tool surfaced it is describing the failure case in Google's own words.

Volume is not the rule, but it is the evidence

Google publishes no number, and anyone who tells you the safe limit is four posts a month invented it. What volume does is make your purpose legible. Nobody can read your intent, but a pattern is readable at a glance: forty near-identical posts on subjects the business has never touched answers the "why" question without anyone having to ask it.

Which is the practical reason to publish less than you technically could, and to keep every piece tied to work you actually do.

A test you can run before you publish

Ask whether the piece exists because a customer asks that question, or because a tool said the phrase had volume. Ask whether anyone at the business could defend every claim in it out loud. Then ask whether you would be comfortable if a customer knew exactly how it was made.

Content that survives those three is fine regardless of what drafted it. Content that fails them was going to age badly whoever typed it, because the thing Google keeps describing is not a technology, it is a motive.

If someone has quoted you a monthly volume of AI-written posts and you want a second opinion on whether it will help or quietly hurt, book a strategy call. The related question of whether AI search will even surface it has a documented answer too, and it is shorter than the pitch.