Chrome is introducing warnings for potentially deceptive or spammy notifications on Android devices. These warnings aim to protect users from unwanted and potentially harmful notifications that might lead to malware downloads, personal information theft, or fraudulent purchases. The feature uses on-device machine learning to analyze notification content, identifying potentially deceptive or spammy messages. When a notification is flagged, users will see a warning and have options to unsubscribe or view the content. The on-device analysis protects user privacy as notification content is not sent to Google. The machine learning model was trained using synthetic data generated by a large language model. This feature is currently available on Android, with potential expansion to other platforms in the future. Chrome already protects against harmful notifications by revoking permissions from abusive sites. Users can also unsubscribe from notifications with one click on Android and manage their notification settings. Chrome's security team is actively working to protect users from online threats.