Nearly everyone who has used the internet has likely encountered alarming pop-ups claiming their device is infected with a virus, urging them to click for tech support or download security software. These “tech support” scams are widespread, and Google is intensifying its fight against them by leveraging cutting-edge artificial intelligence (AI) technology.
AI-Powered Scam Detection Across Google Platforms
Google has integrated a variant of its Gemini AI model, called Gemini Nano, directly on users’ devices to detect and warn against tech support scams in real time. This on-device large language model is now part of Chrome’s Enhanced Protection mode on desktop, enabling the browser to analyze webpages instantly and identify potential threats-even those that use sophisticated evasion tactics like “cloaking,” where scammers hide their true content from traditional detection tools.
On Android, Chrome has introduced AI-powered alerts for suspicious website notifications, giving users the option to unsubscribe automatically from dubious sources. Google Messages and Phone by Google also feature on-device scam detection for calls and texts, scanning for scam-like behavior and alerting users while preserving privacy by keeping all processing local to the device.
The Growing Challenge of AI-Enabled Scams
The rise of AI has unfortunately empowered scammers to generate large volumes of convincing fake content, lowering barriers to fraud that can steal money or personal data. The Global Anti-Scam Alliance reports that consumers worldwide lost over $1 trillion to scams last year alone. In response, Google is deploying AI not only to detect scams more effectively but also to block fraudulent search results and websites. Since introducing AI-powered anti-scam systems three years ago, Google has increased its capacity to block scammy pages by 20 times, significantly reducing scams impersonating official services by over 70% in 2024.
Enhanced Language Understanding and Real-Time Protection
Google’s senior director of engineering for Search, Phiroze Parakh, explained that combating scammers is an ongoing “evolution game” as fraudsters adapt to new defenses. Recent AI advances have improved Google’s ability to understand language nuances and detect coordinated scam campaigns across multiple languages, allowing faster and more accurate identification of threats.
The on-device Gemini Nano model enhances speed and privacy, providing users with immediate warnings before they access risky sites. This proactive approach aims to stop scams at the point of contact, preventing users from falling victim.
User Empowerment and Privacy
Google’s AI tools also empower users by allowing them to dismiss false positives and maintain control over notifications. All AI processing happens locally on devices, ensuring user data remains private while delivering robust scam protection.
In summary, Google is harnessing sophisticated AI models like Gemini Nano to fight the escalating threat of tech support scams and online fraud across its platforms. By combining real-time detection, enhanced language comprehension, and user-focused privacy, Google aims to safeguard billions of internet users from increasingly sophisticated scams.