Google enlists Gemini to block Android scam texts and fake bank calls

On 14 October, Google announced a new line of defense against mobile scams: an opt-in Enhanced protection mode in Google Messages that runs Gemini models in the cloud to catch fraud in real time. The move is part of a broader Android security update that also targets fake bank calls, phishing attempts during screen sharing, and apps that silently forward calls.

The headline feature is the Gemini-powered scam detection in Messages. When users opt in, the AI analyzes incoming texts for suspicious patterns - think fake delivery notices, impersonation of banks, or urgent payment requests - and flags them before they do damage. Because the models run in the cloud, they can adapt to new scam tactics faster than static on-device filters.

Google is also closing two loopholes that scammers have exploited for years. Chrome will hide login fields during screen sharing, so a fraudster on a video call can't watch you type your password. And Android will block apps from silently forwarding calls, cutting off a trick used to intercept one-time passcodes from banks.

The timing is no accident. Mobile scams have become a costly and increasingly common threat, with fraudsters using social engineering to trick users into handing over money or credentials. By putting AI at the center of its defense, Google is betting that the same technology that powers its chatbots can also act as a real-time shield for Android's massive global user base.

The opt-in design is worth noting. Enhanced protection requires users to switch it on, which means Google is balancing security against privacy concerns. Cloud-based AI processing means messages are sent to Google's servers for analysis, a trade-off that privacy-conscious users will need to weigh. Google's decision to make it opt-in rather than default suggests it expects some resistance.

For the broader industry, this is a signal. Other mobile platforms have their own protections, but Google's cloud-based approach is different: it can learn from a much larger pool of scam attempts across millions of devices. That could create a competitive gap in scam detection, especially for users who receive a high volume of SMS phishing attempts.

Regulators are watching too. Telecom fraud and robocalls have drawn scrutiny from government agencies, and carriers have been forced to adopt new authentication protocols. Google's AI-powered approach could set a new standard for what mobile platforms are expected to do, potentially pushing regulators to demand similar protections from other players.

For executives and product leaders, the takeaway is clear: AI is no longer just a feature - it's becoming the default defense layer for consumer trust. The companies that deploy AI to protect users, not just to sell them things, will win the next wave of loyalty. Google's move puts pressure on every mobile platform, bank, and fintech app to answer a simple question: what are you doing to stop the scam before it reaches the user?