Several SEO platforms and tools have reported that Google is returning altered or incorrect results when it detects that the query comes from automated crawling tools. This issue could be part of the search engine’s efforts to limit large-scale data collection by AI systems.
It is important to clarify that this “poisoning” does not affect searches performed by human users on Google, but rather the automated queries made by analytics platforms. According to data shared by digital marketing and SEO platform Ahrefs, the scale of the distortion varies by data provider, with between 5% and 20% of search results affected.
Meanwhile, Nozzle recorded a spike in which as much as 43.5% of the SERPs analyzed in its samples showed signs of poisoning, while platforms such as SEOmonitor and DemandSphere have estimated the error rate at between 6% and 11%.
So far, Google has not commented on the matter.
As Semrush, a brand visibility management platform, stated in a statement, this “SERP poisoning” is affecting data providers in the SEO space and appears in two forms:
One case documented by Semrush showed how a search in Brazil for a baby stroller ended up showing dental equipment products simply because they shared the word “carrinho.”
The leading platforms have taken steps to mitigate the issue. Semrush has stated that its systems are able to detect “clearly invalid” results during collection in order to exclude them from its database. “In addition, we have adjusted our data sources based on their quality and added validation checks that flag potentially inaccurate results and recollect them. We continue refining these processes as search engine behavior evolves,” the platform explains.
For its part, Ahrefs has stated that it already has “several filters to detect and reject anomalous results and request accurate replacements. In addition, we are correcting any previous results that may have been affected.”
Although Google has not issued any statement about this practice, the SEO community seems to agree that the main cause lies in anti-scraping measures. With the rapid development of large language models, hundreds of companies are intensively crawling the web for artificial intelligence data collection. In response, and to protect its infrastructure and ownership of its index, Google may have “poisoned” the information it returns, rendering the quality of content captured by third parties unusable.
When looking at these suspicions, it is hard not to recall Google’s decision last year to disable the &num=100 parameter in its search URLs. This change meant that external tracking services (such as SEO tools or AIs) could no longer extract the top 100 results in a single query, forcing them to make ten separate requests for ten results each in order to obtain the same volume of data.
With that precedent and the current alleged SERP poisoning, everything seems to point to the tech giant seeking to reduce the load on its servers caused by large-scale crawling, limit large-scale data collection for AI systems, and steer the analytics industry toward a much more restricted collection model.
If you manage ranking projects, the main recommendation is to stay calm when faced with unexpected ranking changes in your usual reports. Semrush advises comparing the data with Google Search Console, especially if you detect unusual fluctuations involving keywords unrelated to your pages.
Likewise, both Semrush and Ahrefs (support@ahrefs.com) offer users the option of contacting their support teams so they can help review the case.
Photo: generated with ChatGPT Images 2.5
Your email address will not be published. Required fields are marked *
Δ