Osegi
AI·1 min read

Google to Honor Universal IETF AI Training Token in robots.txt

From 2027, Google plans to read a vendor-neutral IETF token in robots.txt that signals whether content may be used for AI training. Google-Extended stays for exceptions.

Illustration of a robots.txt file on a screen with a blocked AI crawler symbol next to a Google logo.
Image: iX Magazin

Google plans to start evaluating a vendor-neutral AI training token in robots.txt from 2027. The token comes from the IETF, the internet standards body, and is meant to let website operators opt out of AI training with a single entry that applies to all AI developers rather than to one company.

Google-Extended, the company's existing control for AI use of web content, will not disappear. According to the announcement, it remains available for exceptions, so site owners can still set rules specifically for Google where they want to.

Why it matters

Until now, publishers who wanted to keep their content out of AI training had to maintain separate robots.txt rules for each crawler or AI vendor. Every new provider meant another line to add and another name to track. A shared, standardized token would reduce that effort and make the signal easier to understand and to check.

The move also gives the IETF approach weight in practice: a standard only helps site operators if the large crawler operators actually honor it, and Google's commitment is a notable step in that direction.

What comes next

Google names 2027 as the starting point, so website operators have time to prepare. Until then, existing rules, including those for Google-Extended, continue to apply. Whether and when other AI companies adopt the IETF token has not been stated in the available information.

Apps worth a look

SUMO: the number puzzle

Tap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.

Tags:googlerobots.txtietfai traininggoogle-extendedweb crawlers

More from this day

See all →

More in: AI

See all →