For several years, carriers have been employing lists to notify users of possible spam and scam calls as they occur. These systems, however, are far from foolproof. A pressing question is: what happens once a user answers the call? At the Google I/O 2024 developer conference on Tuesday, Google showcased a feature it believes will identify potential scams during a call.
This upcoming feature, slated to be integrated into a future version of Android, leverages Gemini Nano, the smallest iteration of Google’s generative AI that operates entirely on-device. The system is designed to detect “conversation patterns commonly associated with scams” in real-time.
Google provided an example of a scam involving someone posing as a “bank representative.” The system is triggered by typical scam tactics such as requesting passwords and gift cards—well-known methods of financial extortion. Despite widespread awareness, many individuals remain susceptible to these scams. Once the system identifies potential scam indicators, it will notify the user that they might be at risk of being deceived.
No specific release date has been announced for this feature, with Google using this preview to demonstrate the future capabilities of Gemini Nano. It is confirmed, however, that the feature will be opt-in.
This opt-in requirement is a positive aspect because, although Gemini Nano operates on-device without uploading data to the cloud, the system still listens to user conversations in real-time. This capability raises privacy concerns among advocates.
On the downside, with the feature being opt-in, some individuals who could benefit most from such protection might never activate it.
