If you have looked at the Page Indexing report in Google Search Console recently and seen a string of pages marked “crawled – currently not indexed”, the instinct is usually to hunt for a technical fault. A broken canonical tag, perhaps, or a robots directive doing something it should not. Sometimes that is the answer. Increasingly, it is not.

Google’s John Mueller and Martin Splitt addressed this directly on the Search Off The Record podcast this month, and the explanation they gave is one that a lot of website owners will not want to hear: in many cases, the issue is not technical at all. It is quality. Specifically, Google is choosing not to index pages because it does not trust the site enough to bother.

What “Crawled – Currently not indexed” actually means

There is an important distinction worth clearing up first, because Search Console reports several different non-indexed statuses and they do not all mean the same thing.

“Discovered – currently not indexed” means Google knows a URL exists but has not yet visited it. “Crawled – currently not indexed” is a different matter entirely. Google has been to the page, read it, assessed it, and made an active decision not to add it to the index. Mueller was asked whether that particular status is often a sign of a quality issue, and his answer was not a soft one. When Google’s systems have serious concerns about the overall quality of a website, they reduce how much of that site they bother crawling and indexing in the first place. If a large share of a site’s pages come back with this status and there is no obvious technical cause, that pattern itself is the message.

Where AI-generated content comes into this

Mueller was careful not to say that AI-generated content is inherently a problem, and that distinction matters. Plenty of pages produced with AI assistance perform perfectly well. The issue he described is narrower and, frankly, more useful than a blanket warning would have been.

The problem arises when a page reads as though anyone could have written it. No unique angle, no evidence of first-hand knowledge, nothing that could not have been produced by typing the same prompt into a dozen different tools. Mueller’s own phrase for this was blunt: “this tells me nothing.” Content that tells a reader nothing gives Google no reason to give that page a place in the index ahead of the thousands of other pages saying the same thing.

This is where his advice takes an interesting turn, one that will resonate with anyone who has ever tried to be objective about their own website. He acknowledged how difficult that objectivity is. It is your site, your business, your effort; the natural instinct is to see it as the best version of itself. His suggestion is to step back and read your own pages as someone with no connection to the business would read them. Would a stranger learn something they could not have found on the top three competitor pages already ranking? If not, that page is a candidate for the kind of quality issue he is describing.

Why this should not be treated as a technical fix

One line from Mueller is worth sitting with for a moment: he explicitly said this is not something you fix from a technical point of view. There is no setting to change, no directive to add, and no sitemap tweak that resolves a trust problem. If a meaningful proportion of a site’s content reads as generic, the fix is editorial, not technical. That means rewriting, consolidating, or in some cases removing pages that add nothing beyond what already exists elsewhere on the web.

For businesses running content programmes at scale, this is the part worth paying attention to. It is entirely possible to have a technically flawless site, fast, clean markup, correct canonicals, sensible internal linking, and still see indexing stall out, because none of that addresses whether the content itself gives Google, or a reader, a reason to care.

What this means in practice

If your Search Console indexing report shows a pattern of pages sitting in “crawled – currently not indexed” with no technical explanation, treat it as a prompt to audit content quality rather than crawl settings. Questions worth asking of any underperforming page:

  • Does this page say anything that could not have been generated by any competitor using the same source material?
  • Is there a first-hand perspective, a specific example, or original data anywhere on the page?
  • Would removing this page and folding its useful parts into a stronger, more comprehensive page do more good than leaving it as is?

Google’s own advice here lines up with what has been true of search quality for a long time: pages that exist to fill a gap in a content calendar, rather than to answer something a reader actually needs answered, tend to struggle to earn a place in the index regardless of how they were produced. AI is not the villain in this story. Content with nothing to say is, and it always has been.