Quick Answer:

Should your business be worried about Google deindexing more pages? Yes, but not because Google is randomly removing valuable business websites from Search. The concern is more specific than that. Google’s own documentation makes clear that crawling, indexing, and serving are not guaranteed, even when a page follows basic Search requirements. That means a page can exist, be discoverable, and still fail to appear in search results if Google does not see enough reason to include it.

For businesses, the practical takeaway is that weak pages may be more vulnerable than they used to be. Thin articles, duplicate pages, stale posts, generic AI-assisted content, weak internal linking, confusing technical signals, and pages with no clear purpose may have a harder time staying visible. Strong content still has a path to Search visibility, but weak content may no longer get a free pass into the index simply because it exists.

Google deindexing more pages sounds like a technical SEO issue, but for businesses, it raises a more practical question: can you still assume that a page will earn visibility simply because you published it? For years, many website owners treated indexing as a routine step in the content process. A blog went live, the sitemap updated, Google crawled the page, and, if the page did not appear quickly enough, someone could open Google Search Console and request indexing. That process was never guaranteed, but it often worked well enough that indexing felt like a normal part of publishing.

That assumption is becoming harder to trust. The bigger story is not that Google suddenly learned how to deindex websites. Google has always had the ability to exclude pages from Search. The more important question is whether Google has become more selective about what deserves to stay in the index, especially as the web fills with thin, repetitive, AI-assisted, and low-value content. In 2025, third-party data and Search Console observations suggested a meaningful wave of index removals, although that evidence should be handled carefully. Google has not released official Google-wide data proving a broad deindexing surge across the entire web.

Still, the pattern is worth taking seriously. If Google is becoming more selective, businesses need to rethink the old publishing model. Publishing content is no longer enough. A page now needs to justify its place in the index through useful information, a clear purpose, original value, strong structure, relevant internal links, and clean technical signals that help search engines understand why the page deserves to be found.

Why This Conversation Is Happening Now

Generative AI did not create low-quality web content. The web already had plenty of thin blog posts, scraped pages, duplicate location pages, affiliate content, and keyword-heavy articles long before AI writing tools became mainstream. What generative AI changed was the scale of the problem. It became much easier to produce large volumes of content quickly, and while some of that content is useful, researched, edited, and original, much of it repeats the same surface-level information already available elsewhere.

Google has acknowledged the risk of scaled, low-value content. Its guidance says generative AI can be useful for research and structure, but using AI tools to generate many pages without adding value may violate Google’s scaled content abuse policy. That distinction matters. Google is not saying that AI-assisted content is automatically bad. It is saying that content created at scale, without meaningful value for users, can become a search-quality problem.

Google had already signaled that this problem was serious. In March 2024, it announced updates aimed at reducing low-quality and unoriginal content in Search. After that rollout, Google said the changes produced 45 percent less low-quality, unoriginal content in search results than before the update. That does not prove that Google launched a formal “AI cleanup mission” in 2025, but it does show that Google was already focused on reducing low-value content before the late-May indexing reports appeared.

What the Evidence Actually Shows

The strongest public data point comes from Indexing Insight, a third-party tool that monitors Google indexing behavior. In a February 2026 study looking back at May 2025, Indexing Insight reported that more than 25 percent of the 2 million page URLs it monitors were actively removed from Google’s index at the end of May 2025. It also reported that some sites saw 15 percent to 75 percent of monitored pages move into a “crawled—previously indexed” state.

That is a striking data point, but it needs context. This was not official Google-wide data. It was a monitored URL set from one third-party platform, and we do not know whether those 2 million URLs represented the broader web. The sample may have included pages that were already more likely to face indexing issues, such as low-performing content, thin articles, older posts, programmatic pages, or URLs being monitored precisely because indexing was already a concern.

Marie Haynes also reviewed Search Console data across multiple sites and reported that many SEOs were seeing drops in indexed pages starting around May 27, 2025. In the cases she reviewed, she often saw increases in pages listed as “Crawled—currently not indexed.” She also noted that many affected pages appeared to be old blog posts, thin pages, or pages that paraphrased existing information without adding original value. Search Engine Roundtable covered similar SEO community reports and emphasized the uncertainty, noting that the pattern could have reflected an indexing algorithm change, a reporting change, a temporary issue, or something else.

So the evidence is not conclusive. The careful claim is this: there is circumstantial, third-party evidence that Google removed a meaningful number of pages from its index beginning in late May 2025. There is not enough public evidence to say Google officially confirmed a web-wide purge.

What Google Has—and Has Not—Said

Google has not published a report saying that it deindexed a certain percentage of the web in 2025. That matters. Without official Google-wide data, the late-May indexing pattern should not be described as a confirmed mass deindexing campaign. The stronger and more defensible point is that third-party observations suggest Google may have become more selective about what it kept indexed during that period.

At the same time, Google’s public comments did not simply frame the late-May reports as an obvious bug. According to Search Engine Roundtable, Google’s John Mueller said he did not see a technical issue based on examples reviewed and that Google’s systems regularly adjust what is crawled and indexed. He also said Google does not index all content and that what it indexes can change over time. That is not a confirmation of a deliberate purge, but it is also not a denial that index selection can change.

That public message fits Google’s official documentation. Google explains Search as a process that includes crawling, indexing, and serving, and it also says that not every page makes it through each stage. This is the heart of the issue. Google has always been selective. What may be changing is how visible that selectivity has become, especially for pages that are thin, redundant, outdated, or not clearly useful.

Why AI-Generated Content Changed the Stakes

AI content is not automatically bad. A well-researched AI-assisted article can be useful when it includes original thinking, accurate information, human editing, examples, structure, and practical value. The problem is not the use of AI by itself. The problem is scaled sameness: thousands of pages that sound polished but repeat the same basic information, cover the same topics, and offer little that helps a reader understand something better.

When many websites publish similar AI-generated explanations of the same topic, Google has less reason to keep every version in the index. If a page does not add anything new, specific, helpful, or trustworthy, why should Google store it, rank it, or surface it? Google’s scaled content abuse policy focuses on large amounts of unoriginal content that provides little or no value, regardless of whether the content is created by automation, humans, or a combination of both. The policy specifically includes the use of generative AI tools to generate many pages without adding value for users.

That is where the AI angle becomes important. The issue is not simply “AI content.” The issue is content that looks mass-produced, shallow, repetitive, or disconnected from real user value. It is reasonable to infer that the AI content boom increased the amount of low-value content Google has to evaluate. In response, Google has stronger incentives to treat indexing as a quality filter, not merely a discovery process. That inference is not a confirmed Google statement, but it fits the evidence and Google’s public policy direction.

Why “Request Indexing” May Not Work Like It Used To

Many site owners have used Google Search Console’s “Request Indexing” feature as a practical fix. A page was not indexed, so they asked Google to index it. Sometimes that worked, especially when the issue was discovery, a new page, a recent update, or a technical problem that had just been fixed.

But that button was never a guarantee. Google’s recrawling guidance says requesting a crawl does not guarantee that a page will appear in search results immediately or at all. Google also says its systems prioritize the fast inclusion of high-quality, useful content. That distinction is important because it changes how businesses should interpret indexing issues.

If Google has not discovered or crawled a page yet, requesting indexing may still help. If you fixed a technical blocker, removed an accidental noindex tag, corrected a redirect, or improved crawl access, requesting indexing may still be worth doing. But if Google has already crawled the page and placed it in “Crawled—currently not indexed,” the issue may not be discovery. Google has already seen the page. The problem may be that Google does not see enough value in adding it to the index at that time.

In that case, resubmitting the same page over and over is unlikely to solve the underlying problem. The page may need better content, stronger internal links, clearer technical signals, more original value, or consolidation with related pages. The question is no longer only, “Did Google see this page?” The more useful question is, “After seeing this page, why would Google choose to include it?”

What This Means for AI Visibility

This is not just an SEO issue. It also affects AI visibility. Google’s guidance for AI features says the same foundational SEO best practices remain relevant for AI Overviews and AI Mode. It also says that, to be eligible as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to appear in Google Search with a snippet.

That makes indexing a gateway. If a page is not indexed, it has little chance of showing up in traditional Google results, and it is also unlikely to be surfaced as a supporting link in Google’s AI search experiences. AI visibility starts with discoverability, but it does not end there. A page also needs to be useful, understandable, credible, and distinct enough to support a search experience that may summarize information, cite sources, and guide users through follow-up questions.

For businesses, this changes how content libraries should be managed. Old posts cannot simply sit forever with no review. Thin posts should not be published just to cover keywords. AI-assisted articles should not all sound like generic summaries of the same public information. If indexing has to be earned, then AI visibility has to be earned too.

What Businesses Should Do Next

The right response is not to delete half your website. A blind content purge can create new problems, especially if pages still have backlinks, internal-link value, historical relevance, or a role in sales enablement. A better response is to start with an indexing and content audit.

Use Google Search Console to identify which important pages are indexed, which are not, and which have moved into statuses such as “Crawled—currently not indexed” or “Discovered—currently not indexed.” Focus first on pages that matter to your business: service pages, cornerstone blogs, event pages, product pages, resource pages, and strong educational content. A low-value tag page or outdated archive may not matter much. A service page or high-quality resource that is not indexed deserves closer attention.

Then ask whether each important page deserves to be indexed. A page is more likely to earn that place when it answers a clear user need, adds something distinct, reflects current information, links naturally to related content, and gives Google a clean technical path to crawl and understand it. A practical cleanup plan may include refreshing valuable but outdated articles, expanding thin pages that still serve a real search need, merging overlapping posts into stronger cornerstone resources, redirecting obsolete pages to better current resources, improving internal links, adding direct answers near the top of articles, clarifying headings, removing accidental noindex tags, and reviewing AI-assisted content for originality, accuracy, and usefulness.

The goal is not to chase Google with constant cosmetic updates. The goal is to make each important page useful enough to justify its place in the index. That requires content judgment, not just technical SEO.

Conclusion

Google deindexing more pages should not send businesses into panic mode, but it should get their attention. The evidence is not official Google-wide proof of a massive web purge. The strongest evidence comes from third-party monitoring, Search Console observations, SEO community reports, and Google comments suggesting that changes in crawling and indexing are part of normal system behavior. Still, the pattern points toward a practical conclusion: Google appears to be more selective about what stays indexed.

That shift makes sense. The web is producing more content than ever, and generative AI has made it easier to publish large volumes of content quickly. Google still needs high-quality web content. It does not need endless low-value pages that repeat what already exists. For businesses, the lesson is straightforward: publishing is no longer enough. Indexing has to be earned.

That means content needs to be useful, distinct, well structured, current, and technically clean. It also means older content libraries need regular attention. Some pages should be refreshed. Some should be merged. Some should be redirected. Some may not need to exist at all. The businesses that adapt will not just protect their SEO; they will also be better prepared for AI-driven search, where visibility depends on whether content can be found, understood, trusted, and surfaced.

Google deindexing more pages may sound like a threat, but it can also be read as a useful reminder. The index is not a storage locker for everything a business publishes. It is becoming a stronger quality gate for what deserves to be found.

Sources:

• In-Depth Guide to How Google Search Works | developers.google.com
• Google Search’s Guidance on Generative AI Content on Your Website | developers.google.com
• Google Search: New Updates to Address Spam and Low-Quality Results | blog.google
• The May 2025 Google Indexing Purge | indexinginsight.com
• An Interesting Look at Which Pages Google Started Deindexing in Late May, 2025 | mariehaynes.com
• Google Indexing Less Since Late May 2025? | seroundtable.com
• Spam Policies for Google Web Search | developers.google.com
• Ask Google to Recrawl Your Website | developers.google.com
• AI Features and Your Website | developers.google.com