Google rolled out its June Android feature drop this week, delivering an AI-powered scam detection tool to Pixel Phones that screens callers before they can say a word.
The tool analyzes incoming calls in real time, identifying potential scammers based on behavioral patterns and flagging suspicious calls for the user before the conversation begins.
What the Feature Does
The system works at the network and on-device level, cross-referencing caller behavior against known scam patterns without routing audio to Google’s servers, according to Google.
That distinction matters: on-device processing means the call content stays on the phone, addressing a core privacy concern around AI call-monitoring tools.
Still, the feature goes further than existing spam-call filters, which typically rely on databases of known bad numbers. This tool evaluates caller behavior in real time, even from numbers that haven’t been flagged before.
No Equivalent on iPhone
Apple’s iOS Does Not offer a comparable real-time behavioral screening tool for incoming calls.
Apple does allow third-party call-blocking apps on the App Store, but those depend on static number databases rather than live AI analysis of caller behavior.
By contrast, Google’s Pixel integration runs the detection natively, with no third-party app required.
Part of a Broader Scam-Fighting Push
Google has moved aggressively on scam and fraud detection across its product line over the past 18 months.
The company introduced scam detection for phone calls in the Phone app for Pixel Devices in late 2024, initially in the United States, using on-device AI to warn users mid-call if a conversation showed signs of a social engineering attempt — a manipulation tactic where callers impersonate trusted figures to extract money or personal data.
The June drop extends that capability, moving the detection earlier in the call flow so users receive a warning before they even pick up.
Why the Timing Matters
Phone scams cost Americans $25.4 billion in 2023, according to the Federal Trade Commission. Losses from imposter scams — where callers pose as government officials, bank representatives, or family members — accounted for the largest share.
Voice-cloning technology, which lets fraudsters replicate a known person’s voice using a short audio sample, has made those impersonation attempts harder to detect by ear alone.
As a result, the industry push has shifted toward screening calls before the user engages, rather than training people to recognize fraud mid-conversation.
Availability
Google confirmed the scam detection feature rolls out to Pixel 6 and later devices running the June 2025 Android security update.
The feature launches first in the United States in English, with broader language and regional support listed as forthcoming in Google’s release notes.
Pixel phones run a version of Android built and updated directly by Google, giving the company the ability to push AI features to those devices faster than Android manufacturers who rely on Google’s base software and add their own software layers on top.
