On October 23, 2026, 9to5Google reported seeing Google Messages’ optional Enhanced Protection in beta build 20261020_00_RC00. The reported mode combines Gemini analysis of conversations with unknown senders, checks of unfamiliar links and redesigned link warnings.

Google announced the mode on October 14. NeoTeo covered that announcement in its earlier report on the planned Google Messages protection. The beta report describes three protections, each with its own device requirements.

Three protections target different scam signals

Google says Enhanced Protection adds cloud-based Gemini analysis of conversations with unknown senders. The model can assess the context of a conversation, rather than treating an unfamiliar link as the only signal.

A separate feature uses Gemini Nano on the device to check unfamiliar links against real-time web signals before someone taps. Redesigned warnings also give people more context about links sent by numbers they haven’t saved in their contacts. These warnings have a broader stated device threshold than Gemini Nano link checks.

Google Messages already offers Standard protection. Google says it uses on-device machine learning to identify known spam patterns. Suspected spam or scam messages may be blocked or hidden in “Spam & blocked,” and a URL may be sent to Google for link-safety verification.

Device requirements depend on the feature

Google lists different eligible devices for each part of Enhanced Protection:

  • Cloud conversation analysis: Pixel 6 and newer; Samsung Galaxy S24 series and newer; Galaxy Z Flip 6 and newer; Galaxy Z Fold 6 and newer; and other Android devices with AICore.
  • Gemini Nano link checks: Pixel 10 and newer; Samsung Galaxy S26 series; Galaxy Z Flip 8; Galaxy Z Fold 8; and other selected devices.
  • Redesigned link warnings: Android 6 and newer, according to Google’s footnote.

A phone that meets the listed threshold for cloud analysis doesn’t necessarily meet the separate requirement for Gemini Nano link checks. The warning feature has its own Android version threshold.

Google describes separate ways message data is handled

Google says Enhanced Protection’s cloud analysis is encrypted and processed in Private AI Compute, a secure, dedicated environment that Google says it cannot access. This description applies to the enhanced analysis of conversations.

Choosing to report a message as spam is a separate action. Google says that a user-initiated report sends the message and its contents to Google for analysis. For Standard protection, Google says machine-learning detection runs on the device, while a message URL may be uploaded for a safety check.

Google Messages can filter suspected scams

Google says suspected spam or scam messages may be blocked or hidden, and acknowledges that people can miss red flags. Its description of these protections does not promise that every scam will be detected.