Security Flaw in Chrome AI Assistant Exposes Privacy Risks

Researchers from Palo Alto Networks identified a serious vulnerability in Gemini Live, an AI assistant integrated within the Chrome browser. The flaw allowed attackers to abuse browser extension APIs to inject malicious scripts into the AI assistant interface. By exploiting this weakness, attackers could potentially access sensitive information such as webcam feeds, microphone recordings, screenshots, or local files without the user’s explicit consent.

This discovery highlights the growing cybersecurity risks associated with integrating powerful AI assistants directly into browsers and operating systems. While such assistants provide convenience and productivity features, they also expand the attack surface if not carefully designed. Google has since released a patch addressing the issue, but experts warn that AI-enabled applications must be carefully audited to prevent similar vulnerabilities in the future.

05-03-2026