Google Search is implementing a new method for handling organic result links by rewriting them to use a redirect format with google.com/goto?url=..., rather than exposing the destination URL directly in the page's HTML. When users click a search result, Google redirects them to the actual page through this encoded link. The encoding is proprietary and does not correspond to a simple base64 representation, appearing instead as an opaque reference to Google's index.

This update has become widespread since late August 2026, particularly for users who are logged out or browsing in private mode. While it may still be experimental, it is no longer limited to a small subset of search results pages. This new format differs from Google's previous redirect wrapper, google.com/url?q=[destination], which included the target URL in a readable query string.

With the new google.com/goto format, the href attribute points to /goto, not the final destination, and the destination URL must be retrieved from the HTTP Location header via a separate request. This approach prevents offline decoding of the URL and requires scrapers to make additional requests to Google to resolve each link.

Google continues to display copies of the destination URLs on the page to render elements like domain names and favicons, but these are separate from the redirect mechanism.

This change is part of Google's broader efforts to deter automated scraping of search engine results pages (SERPs), particularly from AI crawlers and SEO scrapers that collect URLs in bulk to build independent indexes. By requiring an extra request to resolve each link, Google increases the time and resource cost for scrapers and generates detectable traffic patterns that signal automated activity.

This move complements previous measures such as removing the &num=100 parameter, which limited the number of results per page, and enhancing BotGuard/SearchGuard protections.

The technology company Autom.dev first observed the new goto links on a small percentage of SERPs and has since noted their consistent presence in logged-out and private browsing sessions. Autom.dev has updated its Google Search API pipeline to handle these redirects by reading the Location header without following the redirect, ensuring that clients continue to receive usable destination URLs without requiring integration changes.

As Google continues to adjust its search result presentation, services relying on SERP data will need to adapt to maintain reliable access to destination URLs. Autom.dev remains monitoring these developments and plans to update its tools accordingly.