Blog · Guide

How to Spot AI-Generated Fake Google Reviews

The old tells, broken grammar and obvious repetition, are gone. What gives away a generated review now is what it fails to contain rather than how it reads.

For years, fake reviews announced themselves. Odd phrasing, mangled grammar, the same sentence pasted across a dozen businesses. Those tells are largely gone. A generated review today reads fluently, varies its wording, and sounds like a reasonable person had a mediocre experience.

What has not changed is that the writer was never there. That is still detectable, and it is still the basis for removal.

What generated reviews cannot fake

  • Specifics that only a customer would know. Generated complaints stay at a level of detail that could apply to any business in your category. Real customers name the person who helped them, the day, the thing they bought, the room they sat in.
  • Details that match your actual operation. A review complaining about your waiting room when you do house calls, or about a service you do not offer, was written by someone working from a category, not a memory.
  • A plausible account history. Look at the reviewer's profile. Accounts posting reviews for businesses spread across several states within a few days, or with a single review and no other activity, are the pattern worth documenting.
  • Timing that makes sense. Several reviews arriving within a narrow window, especially after a dispute, is coordination regardless of how naturally each one reads.

Be careful with detectors. Tools that claim to identify AI-written text are unreliable and produce false positives on ordinary writing. A detector score is not evidence. The account history and the absence of verifiable specifics are.

Why fluency does not protect them

Google's policies do not turn on writing quality. A review violates the rules if it does not reflect a genuine experience with the business, if it comes from someone with a conflict of interest, or if it is part of coordinated activity. A well-written fake breaks the same rules as a clumsy one. The case rests on the reviewer and the pattern rather than on the prose.

This is also why reporting a generated review as "fake" through the self-service tool rarely works. Read in isolation, it looks like a competent negative review, because it was built to. The evidence lives in the account and the timing, which the report needs to present.

What to do

Screenshot the review and the reviewer's profile, including their other reviews, before anything is edited or deleted. Note what the review claims and what about it does not match your business. If several arrived together, record the sequence. Do not reply publicly accusing the reviewer of being fake, since that reads as defensive to customers and does nothing for the case.

What we do

We document the account and pattern evidence, identify the specific policy the review breaks, and submit the case through Google's official channel. Every removal goes through Google's own approval. It helps to report while the reviewer's account history is still visible, since that is where the evidence lives. The assessment is free, and you only pay if the review comes down.

Have a Review You Want Removed?

The assessment is free and takes less than a minute. You pay only if the review is removed. We'll give you an honest read within one business day.

Submit for Free Assessment