Google is reportedly testing a version of its search interface that hides the “I’m Feeling Lucky” button when some users begin entering a query. The observation adds to a broader discussion about how Google presents AI features near the search box. It does not establish that the shortcut has been permanently removed or that users are being universally redirected into AI Mode.
Search Engine Roundtable reported the test on October 1, 2026. Barry Schwartz attributed the initial observation to Khushal Bherwani and said he could replicate it. The report compares versions with and without the button, while explicitly noting that AI buttons shown beside the unopened search box were already observed.
Separate the new observation from earlier AI tests
The same report refers to an earlier experiment involving an AI Mode replacement. Separately, an August report on homepage AI buttons described a small desktop test that Google’s Robby Stein confirmed. That earlier confirmation should not be treated as confirmation of the October behaviour.
The available evidence is therefore specific: an observed interface variant, reproduced by a second observer. It does not provide the test’s audience size, duration or eventual release decision.
The entry point can shape the search journey
In our analysis, interface choices deserve attention because they influence what users notice before they submit a query. A visible shortcut can invite one action; a prominent conversational option can invite another. Whether a particular change produces that effect requires behavioural evidence.
For publishers, the meaningful question is what happens after the entry point. Does a user reach a conventional results page, an AI response or a longer research interaction? Those experiences can expose different sources and offer different reasons to visit them. A disappearing button alone does not reveal the resulting distribution of traffic.
SEO needs outcomes as well as screenshots
A team assessing this test should record which interface appeared, on which device and in which market, and what happened after the query was submitted. It can then compare relevant acquisition outcomes over time. That sequence helps connect an interface observation with a business question without assuming causation.
Rankings still matter, but they are only one part of the path to a visit. Visibility within a search experience, visits to the site and subsequent actions should be interpreted separately. A change in one measure does not automatically establish a change in the others.
Keep the content response grounded
The test offers no evidence of a new ranking factor or a required optimisation technique. Rewriting pages merely because a homepage button changed would have little basis in the reported facts. A more useful response is to review how the publication serves the queries that matter to its readers and whether its acquisition mix is changing.
The significance of the observation is its position at the start of the journey. Google is experimenting with how users enter search, while AI options are already part of that interface context. For SEO teams, the task is to watch what those entry points lead to and measure the outcomes, rather than infer a completed transition from one UI variant.