Is Google purposefully serving up false data to AI scrapers?
According to a new report from Nozzle, a search engine results page (SERP) monitoring firm, Google may be deliberately manipulating search results that are visible to rank-tracking tools and other automated services.
SERPs are the pages Google displays in response to a search query. Companies often pay service providers like Nozzle to monitor these results and track how their websites rank for targeted keywords.
While monitoring Google SERPs, Nozzle discovered evidence that the search giant appeared to be deliberately poisoning results pages accessed by rank trackers. Google has little apparent reason to serve misleading data specifically to these tools, suggesting that legitimate SEO services may be caught in the crossfire of Google’s efforts to disrupt data scraping by AI bots.
For its report, Nozzle analyzed 27.5 million SERPs and found that more than 660,000 keywords were associated with poisoned results pages. According to the report, approximately 8 percent of SERPs were poisoned. The altered results appeared in both desktop and mobile versions of Google Search.
What does a poisoned SERP look like? According to Nozzle, it looks much like a normal search results page. However, links that typically rank well in legitimate results may be replaced with less relevant or less accurate results, often from social media platforms.
Google has been battling third-party AI scrapers for some time. These automated bots collect Google’s search data, potentially to power competing AI services or reverse-engineer the company’s search algorithms. By poisoning search results, Google could undermine the value of the data collected by these bots.
The problem is that legitimate rank trackers and other search engine optimization (SEO) tools may also be receiving this misleading data.
It’s also possible that Google’s efforts are intended, at least in part, to target rank trackers themselves. Black-hat and gray-hat SEO techniques have evolved in recent years, with some practitioners using AI to manipulate search rankings. Google may have an incentive to make automated tracking and analysis more difficult for those attempting to game its search results.
However, Google could simply block requests from AI scrapers and other automated tools, as it has done in the past. Over the past month, Nozzle’s findings suggest Google may have shifted from blocking certain requests to serving misleading results instead.
If so, that would point to a more active effort to poison the data available to AI scrapers.
