Google Messages Adds AI Scam Protection That Sends Texts to the Cloud

Google Messages is stepping up its fight against scams with a new optional feature called Enhanced protection. When you turn it on, the app will send information from conversations with unknown senders to Google's cloud, where powerful AI models analyze the context to spot scams that unfold over multiple messages. This is different from the existing on-device scam detection, which stays on your phone. The idea is to catch tricky scams that don't look suspicious at first glance.

So why should you care? Scams are getting smarter, and they cost people real money and stress. Enhanced protection could help you avoid falling for a con that seems legitimate. But there's a privacy catch: Google hasn't said exactly what data is sent to the cloud. Is it the full message? Just parts? How long is it stored? These details matter because your conversations are personal. Google assures that the analysis happens in a secure environment called Private AI Compute, where data is encrypted and even Google can't access it. But without full transparency, it's hard to know how much of your privacy you're trading for security.

The feature will roll out over the coming weeks, but only on certain devices. You'll need a newer phone like a Pixel 10 or Samsung Galaxy S26, and you must enable it yourself. Google also hasn't said which countries will get it first. For now, it's a voluntary step you can take to add an extra layer of scam protection—if you're comfortable with the cloud-based approach.