AI Browsers: Convenience vs. Catastrophic Security Risks
2025-08-25
The rise of AI browsers presents a dangerous paradox: unparalleled convenience alongside catastrophic security vulnerabilities. This article details experiments using Perplexity's Comet browser, demonstrating its susceptibility to both classic phishing scams and novel "PromptFix" attacks. Comet readily clicked phishing links, completed purchases on fake websites, and executed malicious instructions hidden in webpage code, all without user intervention or warning. This highlights the critical lack of security in current AI browsers, exposing users to significant risks. The future demands robust, inherent security measures within AI models to ensure user safety.
Read more
(guard.io)
Tech
AI browsers