Introduction
Google has unveiled two AI-powered scam detection tools for Android devices, enhancing user security against fraud and conversational scams. These features, Scam Detection for Messages and Scam Detection for Calls, were announced in a Google blog post on March 4, 2025.
By leveraging advanced AI models, Google aims to go beyond traditional spam protection, offering real-time alerts while maintaining user privacy.
How Google’s AI-Powered Scam Detection Works
These new tools are designed to identify suspicious patterns in messages and calls, preventing users from falling victim to scams that manipulate them into sharing sensitive data or transferring funds.
Scam Detection for Google Messages
The first feature, Scam Detection in Google Messages, identifies potential scam messages in real time, even after an initial message has been received. Key aspects include:
- Detection of Suspicious Patterns: AI scans SMS, MMS, and RCS messages from unknown senders.
- Real-Time Alerts: Users receive instant warnings about potentially fraudulent messages.
- On-Device Processing: To protect privacy, all message scanning occurs on the user’s device.
- User Control: Users can dismiss, report, or block suspected scam messages.
- Global Rollout: Currently available in English in the US, UK, and Canada, with plans for expansion.
Scam Detection for Calls
The second feature, Scam Detection for calls, is designed to identify fraudulent phone calls before they deceive users.
- On-Device Call Analysis: AI processes call audio in real-time without storing or sharing conversations.
- Privacy-First Approach: No recording, storage, or transcription of the conversation.
- Beep Notification: If enabled, a beep will notify participants about scam detection.
- Limited to Non-Contacts: Only applies to calls from unknown numbers.
- Beta Testing: Initially released for Pixel phone users in the US in late 2024.
Google Cloud AI Protection: Strengthening AI Security
On March 5, 2025, Google Cloud introduced AI Protection, a security solution to safeguard AI workloads and data across cloud platforms. Key Features:
- AI Inventory Discovery: Identifies and categorizes AI-related assets.
- Threat Management: Protects against prompt injection, jailbreak, data loss, malicious URLs, and offensive content.
- Integration with Google Security Command Center (SCC): Strengthens AI security measures.
- Expert Consultation: Google’s Mandiant Consultants assist organizations in securing AI systems.
