Do AI search crawlers like OAI-SearchBot or GoogleOther read my speculation rules block?

No. AI search crawlers fetch HTML and follow links; they do not maintain a browser-style speculative loading pipeline and do not act on , rel=prefetch, or rel=prerender hints. Use sitemaps, internal links, and CDN caching to surface priority pages to AI bots (Google guidance, 2026).

Should I still implement Speculation Rules?

Yes, but for human users in Chromium. Reported navigation speedups in production are real (Shopify and multiple publishers) and improve Core Web Vitals, which feed into broader retrieval quality signals. Just do not budget for a GEO lift from this work alone.

Will AI Overviews credit my page faster if I add prefetch hints?

No. Google AI Overviews relies on its existing index plus retrieval, not on speculative loading. Freshness comes from sitemap lastmod, IndexNow where supported, and crawl frequency — not hint blocks.

Are there any risks of adding speculation rules?

Yes — phantom analytics events, side effects on stateful pages, and origin load spikes if eagerness is set too high. Guard analytics, prerender only safe URLs, and start with eagerness: "moderate".

What replaces deprecated rel=prerender?

The Speculation Rules API. It supports both prefetch and prerender actions, cross-origin navigation, eagerness tuning, and URL- or selector-based targeting (Chrome for Developers, 2025).

How do I know if my CDN supports speculation rules at the edge?

Check whether your CDN supports the Speculation-Rules HTTP response header (allows speculation rules to be delivered as a JSON file referenced by URL). Cloudflare, Fastly, and Akamai all support arbitrary response headers; the rule file itself is plain JSON.

Prefetch and Prerender Hints for AI Search Crawlers

Prefetch and prerender hints — including the older and tags and the modern Speculation Rules API — are browser-side instructions designed for human users, not crawlers. AI search bots such as OAI-SearchBot and GoogleOther typically ignore these hints, so the practical value lies in shaping origin and CDN behavior so priority pages remain fast and discoverable for both humans and AI retrieval.

TL;DR

Resource hints like rel=prefetch, rel=prerender, and the Speculation Rules API speed up navigation for human users in modern Chromium browsers, but AI search crawlers (OAI-SearchBot, GoogleOther, GPTBot, ClaudeBot, PerplexityBot) do not act on them. Use these hints to optimize human UX and to drive your CDN cache strategy; rely on sitemaps, internal linking, and edge caching — not speculative loading — to surface priority pages to AI crawlers.

What prefetch and prerender hints do

Prefetch and prerender hints tell a browser to start fetching or rendering a page before the user navigates to it. There are three implementations a developer should know about, in roughly chronological order.

— the original mechanism. The browser queues a low-priority background fetch of the listed URL and stores the response in the HTTP cache. It is broadly supported but blocked by Cache-Control: no-store and is not optimized for cross-origin navigation (MDN, 2025).

— deprecated in favor of NoState Prefetch and Speculation Rules. Most browsers no longer perform a full prerender from this tag.

The Speculation Rules API, shipped in stable Chromium since Chrome 121, is the modern replacement. Rules are declared in a