ChatGPT faces security threats from Mac vulnerabilities and phishing scams

Security vulnerabilities and fraudulent schemes targeting ChatGPT users have been identified. A critical flaw in the ChatGPT application for Mac recently allowed malware to access user conversations and browser sessions. OpenAI addressed this issue with an application update on September 25, while Apple announced plans to strengthen macOS permissions to enhance user protection.

Separately, a new phishing scam is targeting users through fraudulent Google search advertisements. These ads direct users to malicious custom GPTs that display fake “Service Availability Notices.” The scam instructs users to visit external sites and execute commands in Windows PowerShell, which results in the installation of malware on the victim's computer. Experts recommend navigating directly to chatgpt.com rather than using sponsored search links to avoid these threats.