Using AI to write for your business without getting burned.

What the policy actually says

Google's spam policies name the offense "scaled content abuse": generating many pages whose main purpose is manipulating rankings, whatever tool produced them. In 2026 Google clarified that the same spam policies govern what its AI answers will cite, so there is no separate, softer ruleset for AI visibility. In practice, content demoted in search drops out of those citations with it.

The practical reading: the risk lives in publishing volume without judgment. It lives in pages that would read the same for any company in any city, social proof you did not earn, and markup describing things a visitor cannot see.

The loop we run instead

We draft with AI because it is fast. Then every page passes a human gate before it ships.

• One page, one query. Every page ships targeting the single search need it must earn.
• A named human author with real proof. If a page would be true of any competitor with the names swapped, it does not ship.
• Publish on the KPI, not the calendar. Google has said publishing frequency is not itself a ranking factor. If there is nothing worth saying, publish nothing.
• Re-measure and revert. Every change ships with a re-read date a few weeks out. Anything the search data says made things worse gets reverted.

The test we apply to every tactic

If a tactic's value depends on Google not noticing it, it is banned. That one sentence removes fake reviews, doorway pages, invented ratings markup, and rented authority in a single stroke, and it is why we can automate the plumbing aggressively while a human stays on the trigger for anything a customer reads.