CyberNotableConfirmed
10.0
Claude Flaw Enables Malicious AI Prompts
Dark Reading·US·2 days ago
A security flaw in the Claude for Chrome extension allows malicious third-party extensions to simulate user interactions, potentially granting unauthorized access to integrated services like Gmail, Google Docs, and Salesforce. The vulnerability highlights risks in browser-based AI integrations where cross-extension communication can bypass intended user-consent workflows.