Google Says Chrome Cut 7 Billion Unwanted Android Notifications Per Day

Google announced that its Chrome browser has managed to reduce unwanted Android notifications by over 7 billion per day in the first quarter of 2026, employing multiple defense mechanisms. The company has been working on a "Swiss cheese" model to tackle this issue, combining several strategies to filter out spammy and fraudulent notifications.

Firstly, Chrome can revoke notification permissions from websites that exhibit abusive behavior, repeatedly trigger warnings, or become inactive. This action effectively cuts off their ability to send alerts to a user's device. Secondly, the browser looks for signs of coordinated behavior across multiple websites. Google uses service-worker activity as a signal to identify such coordination.

Additionally, Google's Firebase Cloud Messaging (FCM) adds another layer of control by rate-limiting sites that send notifications at excessive volumes, with a limit of 1,000 messages per minute for repeat offenders.

Furthermore, Google is now offering users a more straightforward way to turn off notifications, which will be prominently displayed within push notifications. This feature aims to make it easier for users to manage their notification preferences without navigating through extensive settings.

The 7-billion figure represents an accumulation of Google's efforts to improve Chrome's notification defenses over several months, including the incorporation of on-device Machine Learning. This local analysis can scrutinize notification content without transmitting it to Google, thereby enhancing the security of user data.

However, the effectiveness of these measures is contingent upon the hardware capabilities of Android devices. Google has not clarified whether the on-device Machine Learning layer is accessible on budget and older devices to the same extent as newer models. This aspect could influence the consistency of protection across different devices.

Despite these sophisticated defenses, Google acknowledges that the notification channel cannot be entirely risk-free. Notifications might still be deceptive, evading all automated defenses. Therefore, user awareness remains the final safeguard, emphasizing the importance of individual vigilance in discerning legitimate notifications from malicious ones.

Google has not reported any specific instances where these notification defenses have incorrectly identified or restricted content. Nevertheless, the presence of automated systems introduces the possibility of false positives, where legitimate sites might be mistakenly flagged. This could frustrate users who rely on receiving notifications from these sites.

It is also crucial to note that automated security systems, including Google's multi-layered approach, are not infallible. The risk of false positives underscores the necessity for user awareness and caution when dealing with notifications.