Google is actively fighting back against the escalating threat of cyber scams, with their latest weapon being artificial intelligence. Recognizing that scammers are using increasingly sophisticated methods, Google is rolling out two new AI-powered features for Android devices designed to identify and thwart these attacks. These features focus on both text messages and phone calls, offering users a more comprehensive shield against potential fraud.
Scam Detection for Messages
The first of these features enhances the existing spam protection within the Google Messages app. This new scam detection system aims to identify a wider range of threats by employing on-device AI that monitors text conversations in real-time. Even if a message starts off innocently, the AI analyzes the flow of the conversation to detect any suspicious shifts in the conversation.
For example, if a seemingly normal conversation begins to veer into an unexpected area, the AI will flag it as potentially fraudulent, the user will receive an alert, offering the option to dismiss the warning or report the sender and block their number. This proactive approach to detection provides an advantage over security measures that only block or allow messages based on their initial content, providing real-time protection.
This feature is compatible with SMS, MMS, and RCS messages, ensuring users are protected regardless of the messaging format. While scam detection within Google Messages is enabled by default, users retain the option to turn it off at any time. It is also worth noting that this feature focuses solely on conversations with numbers not found in the user’s contact list, increasing privacy.
Google reassures users that their exchanges will remain private. When a suspicious thread is reported, only the sender’s information along with the most recent texts are shared with Google and the user’s mobile carrier.
This detection system has already begun its rollout in ENGLISH across the US, UK, and Canada, with expansion to additional countries planned for the near future.
Scam Detection for Phone Calls
Next up is scam detection during phone calls. This feature, which launched last November as a public beta for some Pixel 9 users, also leverages AI to analyze conversations in real-time and warn users of potential scams. During initial testing, the AI’s performance was compared between Pixel 9 and Pixel 6 phones, with the Pixel 9 devices using the more sophisticated Gemini Nano AI demonstrating superior detection capabilities.
The focus is on the detection of scams during phone calls. This can be triggered by situations where a caller requests payment via gift card to complete a purchase – a common type of fraud. The on-device AI then immediately alerts the user through audio, haptic, and on-screen notifications, warning them of attempted fraud.
The phone call feature ensures user privacy – it will not activate when a user is speaking to someone in their contact list. Audio data isn’t recorded, stored, or sent to Google or other third parties, as the processing occurs entirely on the device. However, this function is set to ‘off’ by standard, meaning that users must enable it. In all cases, users retain the ability to switch off this feature even whilst on a call.
Google’s aggressive action to keep one step ahead of scammers is a welcome move. While it’s disappointing that the phone call scam detection feature is not yet available to all Android devices, it is good to see it expanding in the USA and hopefully it will shortly be available to more users. It is understandable that the most advanced AI models need to be supported by the latest technology.
