ChatGPT Using Google’s Index: SEO Is Back :)))
SEO is back without ever going anywhere :)
ChatGPT used to rely on Bing and its internal search functionality to find sources.
There have been multiple tests over the past two weeks showing that ChatGPT may have silently started using Google search:
This test involved a brand new page, not linked from anywhere on or off the site. The page was submitted to Google index through the Search Console, and ChatGPT was able to “read” the unique text from the page once it got indexed by Google.
Another experiment by Aleyda Solís confirmed the above, but it also showed that ChatGPT used Google’s exact search snippet from a new page AFTER it was indexed.
>>>JOIN US to discuss this topic live and ask your questions on LinkedIn!<<<<
Instead of going to the page and fetching information from it directly, ChatGPT waited for Google to index it first (which is unbelievable at this point, many months after OpenAI announced its own search with quite some fanfare).
So why care?
Well, this move isn’t going to change any solid SEO strategy for sure… Quite obviously:
1. Being indexed by Google is fundamental for LLM visibility/AIO (or however we call it these days)
Both tests show a very heavy reliance of ChatGPT on Google’s index. Given that the second most powerful player in AI (Gemini + AI Mode) is also relying on Google’s index, having your site well indexed by Google is the key fundamental step to being findable by LLM platforms.
As Google has started being aggressive in deindexing pages, it is important to keep an eye on your site’s indexation status to make sure all the important pages are indexed.
“Crawled but not indexed” statuses are becoming more frequent. If your product pages are not indexed, there’s very little chance they will be surfaced by major AI platforms in response to relevant prompts.
When looking at your indexation reports in the Search Console, here are a few tips:
It may not be an issue. All sites will have unindexed pages. I don’t think I have ever seen a website with 100% indexed pages :) In many cases, it is not a problem (or at least not a crucial one). Those could be similar pages, outdated or old pages, or search result pages, etc. Unless it’s an important landing page, or key product pages, or there’s a fast-growing trend of more deindexed pages on your site, there’s likely nothing to worry about.
Internal linking structure should be the first crucial step to indexing more internal URLs. Solutions like “Related products”, “People bought together”, etc, get crawlers to go deeper into the site. Resurfacing more categories and subcategories in the main navigation and breadcrumbs will also help.
Lack of unique content may also be the reason for non-indexation. It often happens to larger, product-listing, database-driven websites.
2. Additional AI-friendly optimization efforts are still important
While indexation is fundamental, visibility in generative AI answers is not that predictable. There’s no solid evidence that Google’s organic rankings play any major role in how often URLs are cited from AI answers in ChatGPT or AI Mode / Gemini.
AI algorithms pick pages that provide clearer and more confident answers. So make sure your page:
Has a clear structure based on the HTML headings
Gives clear answers to relevant questions
Uses FAQ Schema or HowTo Schema (actually, no relevant type of schema would hurt)
Allows access to content with JavaScript disabled
And don’t forget that AI answers won’t just address the prompt directly. It will fan-out in many directions, so creating more pages addressing those journeys helps you appear more in AI answers.


