Google is enhancing Android security with AI-driven scam detection across Messages and Phone, utilizing on-device Gemini models and automated link analysis to identify fraud attempts.
Google is expanding Android scam detection with several security upgrades designed to identify and thwart increasingly complex fraud attempts. The central update introduces an optional Enhanced protection mode inside Google Messages. When enabled, the messaging client connects to high-end Gemini cloud models through a private architecture to gauge a sender’s intentions. Google notes that this system issues warnings even when an attacker tests an unfamiliar tactic.
Alongside cloud processing, the Enhanced safeguard uses on-device Gemini Nano models to inspect suspicious web links that previously escaped detection, including freshly created domains used for malware or phishing. When potential threats arise, redesigned warnings present additional checkpoints, such as clearer explanations, to encourage users to think carefully before tapping. Similar link warnings previously required Android Advanced Protection or were restricted to specific regions.
Other enhancements help identify scams during voice calls and screen sharing sessions. Screen sharing now conceals sensitive data, such as banking details and account passwords, inside Chrome and WebView environments rather than restricting protections to native applications. Google is also rolling out defenses against unauthorized call forwarding. The system alerts users if a perpetrator attempts to quietly redirect calls to another phone number or asks the victim to enter a forwarding code.
At the same time, existing notification scam alerts are rolling out to “hundreds of millions” more devices, according to Google. That expansion covers handsets from Motorola along with Chinese smartphone makers including Honor, OnePlus, and Xiaomi.
Not every scam detection tool is guaranteed to operate on every Android phone. Enhanced Protection in Google Messages requires hardware compatible with AICore, such as the Pixel 6, Galaxy S24, Galaxy Z Fold 6, and Galaxy Z Flip 6. Dedicated Gemini Nano link checks demand newer platforms, including the Pixel 10, Galaxy S26, Galaxy Z Fold 8, and Galaxy Z Flip 8. In contrast, the link checkpoints are more accessible, requiring Android 6 or later when they arrive later in October.
The remaining tools follow staggered release schedules. The screen sharing privacy protections require at least Android 15. The call forwarding defense is scheduled for early December on Pixel phones and “other select devices,” while the expanded notification warnings are expected later this year. These software additions do not remove personal responsibility. Users still need to remain skeptical of incoming text messages and verify suspicious links, though Google’s setup functions as a safety net against more creative scammers.
Apple Watch Series 12
Luiza Mosneagu is Reviews Editor at Droid Tools. A psychologist by training and tech writer by passion, she has tested over 60 Android devices since 2020 - always for a minimum of two weeks before writing a word. Her focus is camera quality, everyday usability, and cutting through marketing claims to tell you what a device is actually like to live with.
A new leak suggests the Huawei Nova 17 series could launch next month with a 200MP main sensor, a 50MP telephoto camera, and a Kirin 8 series chip.
Samsung may open a One UI 9.5 beta soon after test firmware appeared for the Galaxy S26 Ultra, ahead of an expected release alongside the Galaxy S27.
Silicon-carbon batteries push smartphone capacity to 10,000mAh, but full recharge times have grown as phone makers likely prioritize cell longevity.
Google took down an Android TV botnet linked to Videostrong, but researchers warn that infected devices may require replacement.
The next story loads as you reach the end. You can also load it using the button.
Comments & Discussions
Join the conversation! We use Disqus to handle comments. Click the button below to load the comment section.