Google has added a note to its Common Crawlers documentation saying it will support AIPREF, a shared machine-readable signal that lets site owners state a preference about AI training, in early 2027. The note covers the user-agent __usage_ai-training, and it was flagged by the TrendQuotient tracker that watches Google's SEO documentation on November 5, 2026. For publishers who have been asking for a clear, standard way to say no to training use, the update is a meaningful marker, even though the support itself is still months away.
AIPREF is a shared machine-readable training preference signal, which means it is designed to be read by machines rather than argued over by lawyers. Instead of each publisher writing its own terms page and hoping a crawler operator reads it, the signal gives site owners a standard way to express a preference about whether their content should be used for AI training. Google's documentation update does not switch the signal on today. It adds a note saying support for the user-agent __usage_ai-training is coming in early 2027, which puts publishers on notice that the mechanism is on Google's roadmap rather than merely under discussion.
The distinction matters because a note in crawler documentation is how Google typically communicates changes to the people who run websites. It is not a blog post or a press announcement, and it is easy to miss unless a tracker is watching the docs. That is exactly what happened here: the TrendQuotient Google SEO documentation tracker flagged the change on November 5, 2026. For technical SEOs, the takeaway is that the documentation itself is now the source of truth for the timeline, and anyone managing crawler access should be reading these pages as closely as they read algorithm update announcements.
Publishers have spent the last few years in an awkward position. Their content powers the answers that AI systems give, yet the controls for opting out of training have been fragmented, inconsistent, and hard to verify. A shared signal, if it is widely adopted, changes that dynamic by giving every site the same vocabulary for expressing a preference. Google saying it will support AIPREF in early 2027 is significant precisely because Google operates at a scale where its choices set norms. When the largest crawler operator commits to reading a signal, other operators face pressure to do the same, and publishers gain a control they can actually rely on.
The timeline also matters for planning. Early 2027 is close enough that publishers should be thinking now about how they want their training preferences set, rather than treating the question as a distant hypothetical. Editorial teams, legal teams, and technical teams will need to agree on a position before the signal goes live, because a machine-readable preference is only useful if it reflects a deliberate decision. Sites that wait until support arrives may find themselves rushing to configure something they have not properly discussed, while sites that prepare early can switch the signal on the moment Google starts honouring it.
The first step is simply to understand the signal and the user-agent it applies to. The documentation note names __usage_ai-training as the user-agent for the AIPREF preference, so technical SEOs should start by checking how their sites currently handle crawler directives and where a training preference would fit alongside existing controls. This is also a good moment to audit which crawlers are actually visiting a site and what they are doing there, because a preference signal only has value if the site owner knows who is fetching their content in the first place.
The second step is to watch for the follow-through. A documentation note is a promise, not a delivery, and the real test will be whether Google's crawlers honour the signal once early 2027 arrives. Site owners should keep an eye on the Common Crawlers documentation for further updates and be ready to test the behaviour when support lands. For link builders, the broader lesson is that the relationship between publishers and AI platforms is being renegotiated in public, one documentation update at a time, and the sites that pay attention to these signals will be the ones best placed to protect their content while staying visible in search.
Sources: Google Search Central documentation, via TrendQuotient tracker, November 5, 2026.
Linkbaits helps you create content that earns citations and backlinks. Start with the free tools and build your first linkable asset this week.
Start free trial