Google confirms it is “running a small experiment” showing blue verified checkmarks beside official site links for companies like Microsoft, Meta, and Apple
Google is experimenting with a new verification feature in search that should make it easier for users to avoid clicking on fake or fraudulent website links.
Context & Ripple Effects
Google is extending verification cues from communications and advertising into its core search interface. Its earlier Gmail checkmarks for verified brands and verified advertiser-name disclosures established adjacent uses for identity signals.
The search experiment matters because it applies that trust cue at the point where users choose among links, initially for prominent companies including Apple, Meta, and Microsoft.
First-order effects
- Users included in the small experiment can see a blue verification marker beside official company links, giving them an additional signal before visiting a site.
- Google, Apple, Meta, and Microsoft become the immediate test participants; Google can evaluate whether the marker helps distinguish official destinations from lookalikes or fraudulent links.
Second-order effects
- If the signal proves useful, other brands and site owners may seek access to the same verification treatment, making eligibility and verification standards a more consequential part of search presence.
- The experiment could shift more attention toward Google’s own identity indicators, complementing the company’s existing verification signals in Gmail and ad disclosures.
Third-order effects
- Search interfaces may increasingly treat provenance as a visible ranking-adjacent layer: not just which link is relevant, but whether its operator is authenticated.
- That direction would make verification governance a durable platform power, with the practical value depending on whether users understand what a checkmark does—and does not—guarantee.
The trend: Google is bringing identity verification cues across more of its consumer surfaces as impersonation and fraudulent destinations become an interface-level trust problem.