{"id":30800,"date":"2025-07-07T07:49:53","date_gmt":"2025-07-07T13:49:53","guid":{"rendered":"https:\/\/housefresh.com\/?p=30800"},"modified":"2025-07-07T08:38:25","modified_gmt":"2025-07-07T14:38:25","slug":"beware-of-the-google-ai-salesman","status":"publish","type":"post","link":"https:\/\/housefresh.com\/beware-of-the-google-ai-salesman\/","title":{"rendered":"Beware of the Google AI salesman and its cronies"},"content":{"rendered":"\n

Sixteen months have passed since we showed you<\/a> how Google\u2019s algorithm was killing independent websites by favoring big media publishers that were abusing their reputation<\/a> to sell you bad products.<\/p>\n\n\n\n

Since then, Google enforced a new spam guideline called “site reputation abuse<\/a>” by removing<\/a> entire sections of major websites from their search index\u2026 at least for a few months<\/a>.<\/p>\n\n\n\n

While this drama was unfolding, Google rolled out AI Overviews<\/a> under the slogan, \u201cLet Google do the searching for you.\u201d<\/em><\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

But are AI Overviews leading you to the best results on the web? Or is this just a new prime spot for selling you products you will regret buying? And how does this change affect the websites that made Google\u2019s AI possible in the first place?<\/p>\n\n\n\n

To find out, we spent weeks conducting research into air purifiers we reviewed<\/a> (including models that don\u2019t exist) to determine whether Google Search results and AI Overviews are leading consumers to good advice.<\/p>\n\n\n\n

It turns out this rabbit hole is deeper than we thought, with ramifications that span major Reddit communities and could be leading you to scams and defamatory hallucinations. <\/p>\n\n\n\n

I hope you\u2019re ready for a long read because this article is packed with examples, fresh data and way too many screenshots.<\/p>\n\n\n\n

Google AI Overviews are mining “product facts” from press releases, product listings and sponsored reviews<\/h2>\n\n\n\n

Don\u2019t bother asking Google if a product is worth it; it will likely recommend buying whatever you show interest in\u2014even if the product doesn\u2019t exist.<\/h3>\n\n\n\n

We found a problematic pattern after spending hours vetting the sources of pros, cons and overall product recommendations spouted by Google\u2019s AI Overview: most \u2018facts\u2019 are sourced from the manufacturer itself, online shops, sponsored articles and press materials.<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

That could explain why AI Overviews used overly salesy language, regardless of what we were searching for.<\/p>\n\n\n\n

This was the case even when asking about models that had been deemed as \u2018the worst air purifier ever tested<\/a>\u2019 by The Wirecutter or had been labeled as \u201cnot living up to the hype<\/a>\u201d by Consumer Reports.<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Not only that… We also noticed that the AI Overview failed to include the very same cons that their algorithm had no issue highlighting in organic search results:<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

In fact, the overly salesy language would still be there even when we asked specifically about cons:<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

This changed once we switched to requesting \u2018a good reason not to buy\u2019 specific models, showing that Google AI Overviews are capable of pulling out critical reviews. <\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Unfortunately, in many cases, the cons themselves were simply hallucinations, which could be easily confirmed by checking the cited source or the organic search results:<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Funnily enough, Google\u2019s organic search results don\u2019t hallucinate\u2014who would have thought.<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n

\n\n
<\/path><\/svg><\/span>Google’s AI will find ways to sell you products that don’t exist<\/strong><\/strong><\/span><\/div>\n\n\n\n

Our tests showed that Google\u2019s AI Overview was positive even when we asked about air purifier models that were completely made up.<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

As reported by Jalopnik<\/a>, “Google […] now relies on an AI that has no concept of “truth” \u2014 just repeating words it sees used in proximity to each other, like your phone keyboard’s autocomplete or a particularly dumb parrot.”<\/p>\n\n<\/div>\n\n\n

<\/p>\n\n\n\n

The more we searched, the clearer it became that no matter the product, the AI Overviews salesman would always repeat a version of the same script:<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

It all started to make more sense once we visualized our data.<\/p>\n\n\n\n

43.1 % of the facts cited by AIO came from the product manufacturer.<\/strong> This statistic combines the cases where statements were linked directly to pages on the manufacturer\u2019s website (19.4%) with facts sourced from product listings built with manufacturer-provided data on popular retailer websites (23.7%).<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Next, we clicked on every link to assess whether it was relevant to our question and the information presented by the AI Overview. We found that 19.5% of the pages sourced by AIO were irrelevant to the query,<\/strong> with the majority of these irrelevant pages being listed when Google\u2019s AI had constructed an answer to questions about air purifier models that don\u2019t exist.<\/p>\n\n\n\n

We also categorized the type of page the AI Overview was sourcing its facts back to and found that a whooping 38.6% of sources pointed to product listings<\/strong> (on the manufacturer website and online retailers) and PR content <\/strong>(including sponsored reviews, sale announcements, press releases and deals coverage on big media sites). <\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Google\u2019s AI Overviews are regurgitating marketing materials and manufacturer-provided facts, while prominently featuring sponsored product listings on top of their glowing recommendations:<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

These search results are a reflection of Sundar Pichai\u2019s vision for Google. <\/p>\n\n\n\n

When Bloomberg asked<\/a> Pichai about the concerns regarding the lack of separation between Google’s search and advertising divisions, Pichai stated that \u201ccommercial information is information, too.\u201d<\/strong><\/p>\n\n\n\n

You wouldn\u2019t trust a car salesman to tell you what\u2019s wrong with the car they are trying to sell you, would you?<\/p>\n\n\n\n

So next time you are researching a specific product, scroll down past the AI Overview, unless you only want to find out what the advertisers want you to know. <\/p>\n\n\n\n

Or if you are set on using AI Overviews to research products, then be intentional about asking for negatives and always fact-check the output as it is likely to include hallucinations.<\/p>\n\n\n\n

The endless loop of Google searches and advertisements in result pages<\/h2>\n\n\n\n

Google\u2019s leadership has promised advertisers that AIO will offer users \u201cads that answer,\u201d and they are delivering on that promise.<\/h3>\n\n\n\n

In 2024, Google shared its Ads Product Vision for AI Overviews and clearly explained the strategy behind everything we\u2019re discussing today\u2014you can watch the video here:<\/p>\n\n\n\n

\n
<\/div>\n\n\n\n
\n
\n