How to spot AI-generated review sites and fake "best of" lists

AI-generated review sites are one of the main reasons product searches got worse. They publish lists like "The 10 Best Air Fryers of 2026" for hundreds of products, written from product listings and other reviews rather than from using anything, and earn a commission when you click through to buy. Here is how to recognise them quickly, and where to look instead.

Seven signs of a generated review site

1. No original photos. Every image is the manufacturer's product shot. Nobody photographed the product on their counter, in their hand or next to a ruler.
2. No measurements. Real testing produces numbers: how long the battery lasted, how loud it was, how evenly it cooked. Generated reviews repeat the listed specs instead.
3. Pros and cons that could fit anything. Easy to use, sleek design, may be pricey for some.
4. A buy button on every pick. Each item ends with "Check price" or "View on Amazon", and there is no product the author recommends avoiding.
5. The site reviews everything. Air fryers, hiking boots, baby monitors and car batteries, all from the same three "experts".
6. No named, findable author. Or an author whose headshot and bio appear on other sites under different names.
7. Leftovers. Chatbot phrases like As an AI language model, template slots like [Product Name], or the same paragraph repeated across lists. See what is AI slop?

What a real review looks like

• Photos and video of the product in the reviewer's own space
• A short "how we tested" section with a method and a time frame
• Clear losers as well as winners, with reasons
• An author with a history in the subject, findable elsewhere
• Updates when products change or fail over time

Google's own guidance for review content asks for exactly these things, evidence of first-hand use above all, and its ranking systems are designed to reward it.

Where to find real reviews

• Add "reddit" or a forum name to your search to find people describing long-term use.
• Publications with hands-on testing labs that explain their methods.
• Video reviews where you can watch the product being used, not a slideshow of stock images.
• Verified-purchase reviews on the retailer, read from the most critical upward.

Filtering them out

You can exclude known content-farm domains with -site: operators or blocklists, as covered in how to hide AI-generated sites from search results. Many review farms are built and run at scale, so learning to recognise content farms in general helps too.

Kitsuvo, our no-AI web browser, labels mass-produced pages as slop with the evidence it found, such as pasted chatbot text, unfilled template slots and pages built around ad slots, and filters likely AI-built results out of your searches.