Google AI Overviews malfunction: what changes for the SERP
- The May 22nd anomaly: what exactly happened
- Why is this error structurally different from the previous ones?
- Immediate Impact on Italian SMEs' SERP Visibility
- Three operational areas to be secured immediately
- The most important signal: Google doesn't really control its own system
- What no one tells you: AI Overviews as an uncontrollable variable
- Outlook: what to expect in the coming months
On May 22, 2026, Google AI Overviews generated anomalous responses for some queries. Specifically, searching the term “disregard” returned messages typical of a conversational chatbot, not an informative summary. Thus, instead of summarizing web content, the system responded with phrases like “Got it. If you need anything else, just let me know.” Google removed the AI Overview for that query within a few hours.
However, the episode is not isolated. It raises structural questions about the reliability of AI Overviews as a dominant element of the SERP. For Italian SMEs that invest in SEO and search marketing, this type of instability has direct consequences on organic visibility. In fact, when the AI Overview occupies the zero position with incorrect or absent content, traffic to the sites is unpredictably redistributed.
At SHM Studio, we closely monitor the evolution of AI Overviews and their impact on positioning strategies. Therefore, this article analyzes what has changed, the immediate operational implications, and the outlook for the coming months.
The May 22nd anomaly: what exactly happened
On May 22, 2026, several users reported unexpected behavior from Google AI Overviews on X. When searching for the term “disregard,” the engine returned a chatbot-style response. The message read: “Got it. If you need anything else or have a new question later, just let me know!”. No summary, no web source synthesis.
The news was documented by The Verge, which reproduced screenshots of the anomaly. In the afternoon of the same day, Google removed the AI Overview for that query. In its place, the engine showed a list of news articles about the malfunction itself. Thus, the correction happened quickly, but the problem was already visible on a global scale.
This episode isn't the first of its kind. AI Overviews have previously generated incorrect answers, misleading content, or baseless summaries. However, the nature of this specific error is different: the system confused its own role, responding as a conversational assistant rather than an informational aggregator.
Why is this error structurally different from the previous ones?
Previous errors in AI Overviews primarily concerned content accuracy. The system synthesized incorrect information, but maintained the correct format. In this case, however, the model lost its functional identity. It responded as if it were a chatbot, not a search engine.
This suggests a possible contamination between the system's layers. Google is integrating increasingly deep conversational capabilities into its products. As a result, the boundaries between Gemini, AI Overviews, and Google Search are becoming more blurred. According to MIT Technology Review, this convergence is intentional but carries risks of behavioral regression in models.
Therefore, this is not a simple display bug. It is a sign of architectural instability in a system that handles billions of queries per day. For SMEs that rely on organic visibility, this has concrete implications.
Immediate Impact on Italian SMEs' SERP Visibility
AI Overviews occupy the highest position on the SERP for many informational queries. When they work correctly, they already reduce the click-through rate to organic sites. When they malfunction, the effect is even more destabilizing.
In fact, an incorrect or absent AI Overview alters traffic distribution unpredictably. Users may abandon their search, click on different results than usual, or turn to other search engines. For an SME that has optimized its pages to appear in the sources cited by AI Overviews, a system interruption temporarily negates that work.
We of SHM Studio We observe that many Italian companies do not yet have an explicit strategy to manage the variability of AI Overviews. However, this variability is now structural. It is not a question of if it will happen again, but when. Therefore, strategies need to be adapted. SEO consequently.
Three operational areas to be secured immediately
The first area concerns monitoring presence in AI Overviews. Not all queries trigger this element. Furthermore, presence varies by industry, language, and device. It is crucial to map which strategic queries generate AI Overviews and how often your site is cited as a source.
The second area is traffic diversification. Over-reliance on organic traffic from informational queries exposes SMEs to risks of system instability. Consequently, it is appropriate to balance investments between Organic SEO, Google Ads campaigns e LinkedIn campaign.
The third area concerns the structural quality of content. AI Overviews tend to cite pages with clear semantic markup, structured data, and authoritative content. Therefore, investing in SEO copywriting quality and in a correct information architecture remains the strongest defense against the unpredictability of the system.
The most important signal: Google doesn't really control its own system
The most relevant aspect of this episode is not the bug itself. It's the speed at which it manifested and the nature of the fix. Google didn't solve the problem at its root: it simply removed AI Overview for that specific query.
This reactive approach confirms that the AI Overviews supervision system is still flawed. According to research by Gartner, Search engines based on generative AI will face episodes of this kind with increasing frequency in the next 18-24 months. Therefore, the stability that SEO operators took for granted is no longer guaranteed.
For Italian SMEs, this means that relying on a single digital acquisition channel is an increasingly risky strategy. Conversely, an integrated approach that combines digital marketing, SEO and presence across multiple touchpoints offer greater resilience.
What no one is saying: AI Overviews as an uncontrollable variable
In public debate, AI Overviews are often described as a threat to organic traffic. However, there's a less discussed aspect: they are also a variable that no operator can directly control.
Unlike an organic result, which responds to relatively stable ranking signals, an AI Overview can appear, disappear, or change content without notice. Furthermore, the site cited as a source does not necessarily receive traffic proportional to its visibility in the Overview. Therefore, optimizing for AI Overviews is a legitimate goal, but it cannot be the only one.
In SHM Studio We recommend that SMEs treat presence in AI Overviews as a bonus, not a primary KPI. The main strategy must remain rooted in controllable metrics: organic rankings, direct traffic, conversions from digital campaigns and user experience quality on website.
Outlook: what to expect in the coming months
Google will continue to expand AI Overviews. Last year, in 2025, the feature was extended to new markets and languages, including Italian for some query categories. Following that, throughout 2026, coverage increased further. Therefore, episodes like the one on May 22nd will become more frequent, not less.
SMEs investing in AI-ready strategy and in a structured digital presence will be better positioned to absorb these fluctuations. Finally, those who have not yet started a review of their SEO strategy in light of AI Overviews should do so before the phenomenon becomes further consolidated.
To further explore the operational implications for your company, you can consult the SHM Studio Blog o Contact the team for a personalized evaluation.
News Categories
Related articles
Discover other articles that explore similar topics in depth, selected to give you a more complete and stimulating view. Each piece of content is carefully chosen to enrich your experience.