A Flaw in ChatGPT’s Mac App Could Have Let Hackers Grab Sensitive Data

A patched vulnerability in the macOS ChatGPT app could have let an attacker take over the app, access chat history, and potentially leverage trusted integrations like browser sessions or other sensitive apps. For CIOs and technology leaders, this is a reminder that AI assistants are becoming high-trust endpoints with broad privileges, so their security posture now directly affects data protection, insider-risk exposure, and enterprise attack surface. IT organizations should treat AI apps like any other privileged software: tightly govern deployments, monitor vendor patching, and validate least-privilege controls around connected accounts and systems.