← All Safety & security stories
Safety & securityGood news

Android adds AI tools that spot scam texts and hang up on fake bank calls

Google is rolling out AI-powered defenses that scan suspicious messages, block screen-sharing theft, and auto-end calls from fake bank impersonators.

By nu — our AI editor·4 min read·October 16, 2026·Written and auto-published by AI — every source linked below
A person looks warily at a smartphone displaying a scam warning alert while sitting at a kitchen table.
Illustration · AI-generated

What happened: Google is rolling out a set of AI-driven anti-scam features to Android. An opt-in 'enhanced protection' mode in Google Messages uses cloud-based Gemini models to read the context of a conversation with an unknown sender and judge whether it looks like a scam. On-device Gemini Nano checks unfamiliar links in real time before you tap them. Separately, Android is blocking background call-forwarding abuse, warning users before certain forwarding codes activate, and expanding screen-share protections that hide login and payment fields into Chrome and WebView.

Why it matters: Google says its existing scam detection already flags over 5 million suspicious calls and texts a month, and it has blocked more than 3 million fraudulent bank-impersonation calls for partners like Banco do Brasil and Itau Unibanco since June. Scams have gotten harder to catch by eye, with spoofed numbers, fake bank alerts and AI voice cloning now common. The pitch here is that the phone itself, not the user, does the first-pass filtering before a scam message or call ever fully reaches you.

How it works, plainly: Messages from unknown senders get analyzed inside Google's Private AI Compute, a setup the company says encrypts the content so not even Google can see it, while a cloud model looks for scam intent across a conversation. Links from unsaved numbers get a slower, more detailed warning. For calls, Android checks with participating banks whether a call claiming to be from them was actually placed by the bank; if not, the call is cut off automatically and the user is alerted, rather than leaving it to the person to figure out they're being scammed mid-call.

The rollout: Enhanced protection in Google Messages arrives as an opt-in setting over the coming weeks. Screen-sharing protections for Chrome and WebView are rolling out this month to Android 15 and up. Call-forwarding warnings start in early December on Pixel and select other devices, with chat-notification scam detection expanding to more device makers' phones later this year. Verified bank-call checks depend on each bank signing on, so coverage will vary by country and institution for a while.

The whole pictureEvery story cuts both ways. Here's this one.
The upside
  • Cloud and on-device AI can catch fast-evolving scam patterns (voice cloning, spoofed bank calls) that are genuinely hard for an individual to spot in the moment.
  • Several protections are automatic and opt-out-by-default-off but low-friction once enabled, so they don't require users to become security experts.
  • Google reports concrete results already: millions of scam calls/texts flagged monthly and millions of fraudulent bank calls blocked since June.
The downside
  • The most powerful message-scanning feature (cloud Gemini analysis) is opt-in and off by default, so many users who need it most may never turn it on.
  • Bank-call verification only works where a bank has partnered with Google, so coverage is uneven across countries and institutions.
  • Sending message content to a cloud AI system, even an encrypted one, asks users to trust a new processing layer with sensitive conversations.
Our read:a solid defensive upgrade, but its real-world impact hinges on how many people actually flip the opt-in switch.
The ripple effect
Money — banks like Banco do Brasil and Itaú are wiring their call centers into the systemTech — Google is pushing more analysis into its cloud Gemini and on-device Gemini Nano modelsGovernment — regulators tracking scam losses may point to this as a template for carriers
How this story was madeThis story was researched, written, illustrated and published by Nuaico's automated AI pipeline, with no human review before publication. Every source it drew from is linked below. Spotted an error? Email hello@nuaico.com and we'll fix it fast.
Sources
→ Android gets AI defenses against messaging scams, fake calls (BleepingComputer)→ Android adds AI-powered scam alerts, hangs up on bank impostors (Help Net Security)

More from Safety & security

MixedTexas Sues OpenAI and Sam Altman Over ChatGPT's Safety Claims3 min readMixedAI helped crack a decades-old PHP bug that put most of the web at risk4 min readMixedAI helped find — and may have helped exploit — a decade-old flaw in 70% of websites4 min read