"Mobile scams have evolved far beyond basic robocalls and phishing texts, turning into sophisticated operations that use spoofed numbers, fake bank alerts, and AI voice cloning," Google says.
Google Messages gets Gemini-backed cloud analysis
Google announced a new cloud-based protection for Google Messages that uses Gemini-backed analysis to scrutinize complex conversations for signs of fraud. The company said the system analyzes conversations with unknown senders within an encrypted Private AI Compute environment, which Google says protects user privacy at all times. These cloud protections are currently disabled by default; users can choose to activate them.
On-device Gemini Nano and clearer link warnings
In addition to the cloud checks, Android adds on-device checks using Gemini Nano. Google described these local AI model checks as focusing on unfamiliar links, using real-time web signals to assess risk and presenting clearer warnings before users open links from unsaved numbers. Google framed the two-tier approach—cloud analysis for complex cases and local models for link checks—as an attempt to catch threats that its existing Phone and Messages systems may miss.
Audit-ready is a season. It shouldn't be.
Evidence in spreadsheets, controls drifting between audits, frameworks multiplying on flat headcount. Nubivance runs continuous compliance on Rapid7 Cyber GRC - SOC 2, HIPAA, ISO 27001, PCI, CMMC.
End the scrambleCall-forwarding changes and MMI-code warnings
Google also announced system-level changes to blunt call-forwarding abuse. Android will block apps from silently enabling call forwarding in the background and will present warnings before manually entered MMI forwarding codes take effect. Separately, on-device scam detection in chat notifications—previously launched on Pixel devices—will be extended to "hundreds of millions more users" through Android device partners later this year, according to Google.
Verified bank calls expanded; millions of scams already flagged
Google is expanding its verified bank calls system, introduced earlier this year, to include more financial institutions. That system checks whether an incoming call was initiated by a real bank call center or is a spoofing attempt; if verification fails, the call is automatically ended. Google reported that, since June, it has blocked over 3 million fraudulent calls of this kind for Banco do Brasil and Itaú Unibanco alone. Google also cited support for Revolut and other global financial institutions.
Google noted its existing Phone and Messages scam detection already flags more than 5 million suspected scam calls and texts each month, and said the new measures are expected to catch even more threats as they roll out to eligible devices.
What this means for Android users, financial institutions, and app developers
- Android users: Google advised users to upgrade to the latest Android release, keep apps updated, and independently verify caller or sender identities when receiving account-related requests or requests for sensitive information. Users also face a choice: enable the new cloud protections (disabled by default) or rely primarily on on-device checks like Gemini Nano.
- Financial institutions: Banks already participating in the verified bank calls system gain an automated check against call spoofing; Google cited specific results for Banco do Brasil and Itaú Unibanco in blocking fraud since June. Institutions considering enrollment can expect an automated verification gate that ends calls failing the check.
- App developers and device makers: Android's block on silent call-forwarding and the expansion of on-device scam detection to a broad set of devices through partners change the operational constraints for apps that previously could enable forwarding without user consent and for manufacturers integrating notification-based protections.
Google positioned the suite of changes as a response to attacks it described as increasingly sophisticated—ranging from fake bank alerts to AI voice cloning and spoofed numbers—and as an effort to shift some of the detection burden from users to automated systems. The rollout is under the control of users (cloud protections default off) and device makers (on-device features expanding through partners), making the immediate impact dependent on adoption and activation. Android users who want the new protections must update their devices and, if they choose, enable the cloud-based checks.
