Cloudflare signals add pages to OpenAI search
A Cloudflare–OpenAI pilot using network signals from about 1% of Cloudflare's traffic has added hundreds of millions of previously unknown pages to OpenAI's search index, widening the set of sources ChatGPT can draw on.
Cloudflare shared early results from a research pilot with OpenAI, launched in July 2026, that tests whether signals from Cloudflare's network can help OpenAI's search crawler find web content more efficiently.
According to Cloudflare, the pilot drew on a participating cohort of roughly 1% of its network. It has enabled OpenAI to discover and index "hundreds of millions" of new pages. Cloudflare provides the discovery signals. OpenAI then evaluates and ranks the pages on its own, based on how relevant and helpful they are.
Most of the work involves filtering. Cloudflare says automated traffic can make pages that don't exist look interesting: bots repeatedly request missing files, some sites redirect almost any address to their homepage, and some URLs lead to duplicate content. Spotting these patterns lets a crawler skip wasted fetches. Change signals tell crawlers when a page is worth another visit. The companies are also testing page-level flags for potentially harmful content, such as a fake login form planted on an otherwise legitimate site. Early results suggest these flags can help keep such pages out of an index.
Cloudflare acknowledges that finding new pages can increase crawling at first, while those pages are evaluated. It also says a signal about a page does not grant permission to access it, and that the research uses metadata for search and discovery rather than sharing page content for model training.
The post gives no figures on crawl-load reductions or downstream traffic to publishers. Cloudflare says the next phase will refine page selection, test faster change signals and measure what happens after content is discovered, including business outcomes for site owners.