The short answer: no, Google does not penalise content for being written with AI. Its spam policies target scaled content abuse — making many pages primarily to manipulate rankings rather than to help people — and that policy explicitly covers AI-generated, human-written and scraped pages alike. The test is intent and value, not the tool. In May 2026 Google confirmed the same policies apply inside AI Overviews and AI Mode. For a local business, though, the realistic risk is not a penalty: it is publishing forty generic posts that nobody in your town searches for, on a site that still has no prices and no photos of finished work.
Somebody has probably offered to sell you sixty blog posts a month. Somebody else has probably told you Google will destroy your site if you use AI at all. Both are selling something.
Here is what Google's documentation actually says, and what it means for a business that serves customers within about an hour's drive.
What Google's policy actually says
Google's position has been public since early 2023 and has not moved: appropriate use of AI is not against its guidelines. What is against them is content produced primarily to manipulate rankings rather than to help people.
The relevant rule is scaled content abuse, in the spam policies for Google Web Search. Read the named examples and notice what they have in common:
- Using generative AI to create many pages that add no value for users
- Scraping feeds or search results to generate many pages
- Stitching together content from other pages without adding value
- Creating multiple sites to hide the scaled nature of the content
- Making many pages that make little sense to a reader but contain search keywords
Every example is about volume without value. The policy is deliberately method-agnostic: an outsourced writing farm producing 200 thin human-written pages is in scope on exactly the same terms as a script producing 200 AI ones.
In May 2026 Google updated the same documentation to confirm that all of these policies — scaled content abuse, site reputation abuse, doorway pages, cloaking, the rest — apply to what appears in AI Overviews and AI Mode, not just to the traditional blue links. The rules did not change. Their reach did.
The distinction that decides everything
| Google is fine with | Google treats as spam |
|---|---|
| AI drafting a post you plan, fact-check and edit | Publishing dozens of AI posts a day with no review |
| AI helping outline a service page you then fill with real detail | Near-identical pages for each nearby town with the name swapped |
| AI rewriting your own rough notes into clear prose | Pages assembled from other sites' content with nothing added |
| AI turning your job photos and prices into a case study | Pages written for a keyword that make little sense to a reader |
The left column is a drafting tool. The right column is a publishing strategy. Google cares about which one you are running, and so should you — because the right column also does not work, which is the more useful reason to avoid it.
For a local business, the real risk is irrelevance
Here is the part the AI-writing-tool blogs will not tell you.
If you are a plumber in Kamloops, you are not competing in a market with enough search volume to reward forty posts a month. There are maybe a few dozen queries that actually bring you paying work: your service plus your city, "near me" variations, emergency phrasing, cost questions, and a handful of problem-aware searches. That demand is finite. Publishing at volume into it does not create more of it.
So the failure mode is rarely a penalty. It is forty posts nobody searches for, on a site that still does not list prices, still has no photos of finished jobs, and still does not say which neighbourhoods you cover. You paid for content and bought nothing.
The things that actually move a local site are the things a general-purpose model cannot invent for you:
- What you charge, or at least a range
- Photos of work you have actually done, with the job described
- The neighbourhoods, buildings and conditions you know because you work here
- Real answers to the questions people ask on the phone
- Reviews, in your customers' words
A model does not know that half the houses in Brocklehurst have a particular vintage of plumbing, or what you charge for an after-hours call, or that the tricky part of a Sun Peaks job is access in February. That knowledge is the whole competitive advantage. AI is genuinely good at turning it into clean prose. It cannot supply it.
The trap most local sites walk into
There is one pattern worth naming because it is the most commonly recommended and the most likely to cause an actual problem: one near-identical page per surrounding town.
Kamloops, then Chase, then Merritt, then Logan Lake, then Barriere — same page, place name swapped. That is the doorway pattern Google names in its spam policies, and it does not matter whether a person or a model produced it.
The version that works is the same shape, done honestly: build the page only where real demand exists, and make it genuinely about that place — the businesses there, what is different about that market, work you have done in it. That is more effort per page and far fewer pages. It is also the only version that survives.
How to use AI without wrecking anything
The workflow I would actually recommend to an owner:
- Pick the topic from demand, not from a content calendar. One post that answers a question customers ask on the phone beats ten on themes a tool suggested. The content brief generator will give you the structure for a specific topic and city in about a minute.
- Supply the facts yourself. Your prices, your process, your local specifics, your job stories. Give them to the model rather than asking it to invent them.
- Let it draft, then cut. These tools pad. Delete every sentence that would be true for a competitor in another province.
- Check every fact and every number. Models state wrong things confidently, and a wrong price or wrong regulation on your site is worse than no post.
- Publish less than you think. One genuinely useful post a fortnight, kept accurate, beats a daily stream nobody reads.
- Say who wrote it. A named author and a real business behind the page is part of what both Google's quality framework and AI assistants use to judge whether you can be trusted.
That process uses AI hard, and produces content Google's policies have no argument with, because the value is real.
Does any of this change with AI search?
It sharpens it. As more answers get summarised by AI Overviews, AI Mode and assistants, the number of businesses named in an answer shrinks. What gets you named is specific, verifiable, first-hand information — prices, service areas, real detail — which is the exact opposite of generic AI filler.
The businesses that will struggle in AI search are the ones whose entire web presence could have been generated from their category and city. If a model could write your whole site without talking to you, it has nothing distinctive to cite.
None of which makes SEO obsolete — it still works, and the fundamentals for ranking in Kamloops are unchanged. The bar for what counts as useful just went up.
Common questions
Does Google penalise AI-written content? No. Google's guidance permits appropriate use of AI and evaluates content on quality and usefulness rather than how it was produced. What it penalises is scaled content abuse — making many pages primarily to manipulate rankings — which applies equally to AI-generated, human-written and scraped content.
What is scaled content abuse? Google's term for generating many pages primarily to manipulate search rankings rather than to help people. Named examples include using AI to create many pages without adding value, scraping feeds to generate pages, stitching content from other sites, and publishing pages that make little sense to a reader but contain keywords.
Can I use ChatGPT to write my business blog? Yes, as a drafting tool. Supply your own prices, process and local detail, check every fact, cut anything generic, and have a person review it before publishing. The problem is never the tool — it is publishing volume with no review and nothing original in it.
Will Google know my content was written by AI? That is the wrong question to optimise for. Google's policies are about value and intent, not detection, and its guidance does not promise to identify AI text. A page that is genuinely useful is safe regardless; a page that exists only to rank is at risk regardless of who wrote it.
Should I publish a page for every town near me? Only where real demand exists, and only if the page is genuinely about that town. Near-identical pages with the place name swapped are the doorway pattern named in Google's spam policies, whether a person or a model wrote them.
The honest summary
Use AI to draft. Do not use it to manufacture. Google's line is volume without value, and it has held that line consistently for three years now — but for a local business the harder truth is commercial rather than technical: your market is small, your advantage is what you know about it, and no model can supply that for you. Fewer pages, more of yourself in them.
If you want a read on where your site actually stands before adding anything to it, the free website grader covers speed, mobile and on-page structure in about a minute.
Written by Jagatjeet — Jagatjeet (jagatjeet.com) is a web design, local SEO and AI automation studio in Kamloops, British Columbia, serving the Thompson-Okanagan and BC Interior. Published 10 August 2026. Last updated 10 August 2026.