Google announced a significant expansion of its AI-powered scam protection across Android devices this week, bringing real-time call and message fraud detection to more phones, more countries, and more languages.
Starting with the Galaxy S26 series unveiled at Samsung Unpacked, Google’s Scam Detection for phone calls will appear outside the Pixel ecosystem for the first time. The feature, which analyzes speech patterns during live calls for signs of fraud, previously lived exclusively on Google’s own phones.
The expansion significantly widens Android’s safety net. According to Google’s Security Blog, the company is “continuing to work with our partners to bring these industry-leading protections to even more users.”
Scam Detection for calls already operates in the US, Australia, Canada, India, Ireland, and the UK on Pixel devices. Google didn’t specify whether the Samsung expansion would eventually reach those additional markets.
The system runs entirely on the device using Gemini, Google’s on-device AI model. That means your conversation never leaves your phone. Google emphasized in its announcement that “the phone conversation processed by Scam Detection is neither stored on your device, nor shared outside of the device.”
Users control the feature too. It’s turned off by default and never activates during calls with contacts stored in your phonebook.
Hunting ‘pig butchering’ and job scams
Phone calls aren’t the only battlefield. Text message scams have grown increasingly sophisticated, moving beyond obvious Nigerian prince pitches into elaborate long-con operations such as “pig butchering” (romance investment fraud) and fake job offers. These tactics are subtle and build trust over weeks or months.
To fight back, Google is rolling out enhanced protections in Google Messages across more than 20 countries, supporting languages from Arabic to German. On the latest flagship devices like the Pixel 10 and Galaxy S26, Gemini’s on-device AI can now perform a “nuanced analysis” of these complex conversational threats to spot the red flags that humans often miss.
The same Gemini AI powering scam detection is also learning to handle background tasks like ordering groceries or booking rides, and Circle to Search now identifies full outfits rather than just individual items. But the safety features may matter most, as a scam call doesn’t care if you’re technically sophisticated or a first-time smartphone user. It just needs one moment of weakness.
Also read: A surge of fake Gemini chatbots is helping scammers push bogus crypto presales.
